International Journal of Technology 16(1) 82-96 (2025)

International Journal of Technology

[Link]

Research Article

Motion Control and Obstacle Avoidance of Mobile

Robot with Mecanum Wheels
Hassan M. Alwan 1,*, Volkov Andrey Nikolavich 2, Ali Shbani 2, Kochneva Olga Vladmirovna 2

1
Mechanical Engineering Department, University of Technology, Alsinaa street, 10066, Baghdad, Iraq
2
High School of Automation and Robotics, Peter the Great St. Petersburg Polytechnic University, 29,
Polytekhnicheskaya street, St. Petersburg, 195251, Russia
*
Corresponding author: [Link]@[Link]; Tel.:+9647902597660

Abstract: Most current research on motion control and obstacle avoidance for mobile robot (MR) is
primarily focused on static obstacle avoidance, with limited attention to detecting as well as avoiding
moving obstacle. As a result, significant research is still needed to address the problems associated
with movable obstacle avoidance. The difficulty of movable obstacle avoidance is influenced by
several factors related to obstacle, including its size, geometric dimensions, velocity, direction of
movement, and acceleration. Other parameters related to the kinematics of MR also play a significant
role. Therefore, this research aimed to develop a new method of intelligent detection as well as
avoidance of movable and static obstacle. To achieve this purpose, wheeled mobile robot (WMR) was
equipped with six ultrasonic sensors to detect the distance between the robot and obstacle. Moreover,
a new algorithm was developed based on WMR center position (CP), provided by the control system
or distances between the obstacle and WMR detected by sensors. This algorithm determined the
position and velocity of movable obstacle. According to the outputs related to movable obstacle,
WMR avoided collision by altering its path to follow a feasible alternative path, which was planned
based on individual priority. This method was simulated in avoidance of static and movable obstacle
using MATLAB Simulink program. The results obtained during the analysis showed that the
percentage of maximum additional time to avoid the fixed obstacle was 1.15% while movable obstacle
was 0.7% of the total time. In addition, the percentage of maximum additional length to avoid the
fixed obstacle was 5.52% while movable obstacle was 3.58% of the total desired length. The results
were satisfactory compared to previous research, showing that WMR avoided movable obstacle and
returned to the desired path faster than in other investigations.

Keywords: Obstacle avoidance; Path planning; Robot kinematics Trajectory tracking; WMR

1. Introduction
The use of wheeled mobile robot (WMR) is gradually increasing in daily life, industry, and space
applications. However, as mobile robot (MR) become more incorporated into various areas of
everyday life, challenges are also experienced during the implementation of tasks assigned to the
robot, particularly in the aspect of motion control, as well as fixed and movable obstacle avoidance.
Research has devoted significant attention to addressing these challenges over the past decades.
These efforts have successfully resolved many problems, especially those related to fixed obstacle.

This work was supported by University of Technology as well as High School of

Automation and Robotics at Peter the Great Saint Petersburg Polytechnic University

[Link]
Received August 2024; Revised September 2024; Accepted November 2024; Published January 2025
 83
International Journal of Technology 16(1) 82-96 (2025)

Research on avoiding movable obstacle has not achieved the same level of progress because of its
complexity but several significant contributions in this field can be observed. Following the
discussion, various findings have implemented several methods for WMR motion control including
obstacle avoidance, to ensure smooth task execution along the designated path. (Abed et al., 2024)
aimed to review various path planning strategies for MR using different optimization methods
obtained from recently published papers in the last five years. In (Al-Mallah et al., 2022) finding, a
Type-2 fuzzy logic controller built from human experience was used to guide WMR to reach its
predefined destination by controlling linear and angular velocity of the robot, preventing obstacle
collision. Moreover, research by (Zghair et al., 2023) focused on finding short and smooth path in a
static environment based on hybrid algorithms to avoid collision. In their paper (Abdelwahab et al.,
2020) present the Z-number based Fuzzy Logic control for trajectory tracking of differential wheeled
mobile robots. The unique point of their approach lies in the ability to encode constraint and
reliability in multi-input and multi-output rules. The investigation of (Kuo et al., 2018) allocated a
method of obstacle avoidance and focused on curvature constraint for a non–holomonic MR using
curvature–constrained streamlines changing via pure pursuit. A new dynamic algorithm was
proposed in (Wang et al., 2022) to set the initial direction angle and then improve its evaluation
function during implementation of the task to increase the efficiency as well as the flexibility of the
algorithm. Rusdinar et al. (2021) developed MR using fuzzy inference system algorithm for its
movement with a magnetic line sensor. Following this discussion, the research by (Dang and Bui,
2023) used a binary semantic segmentation (FCN – VGG – 16) to extract features from image
captured by monocular camera. This model estimated the position and distance of obstacle in the
robot environment. Based on the improved algorithm, the optimized path planning increased the
performance of the path of MR. (Feng et al., 2021) paper set out to investigate the usefulness of
solving collision avoidance problems with the help of deep reinforcement learning in an unknown
environment, especially in compact spaces, such as a narrow corridor. Their research aims to
determine whether a deep reinforcement learning-based collision avoidance method is superior to
the traditional methods, such as potential field-based methods and dynamic window approach.
Finding by (Wang et al., 2021b) was aimed at developing a review of research progress about local obstacle
avoidance method for WMR under complex unknown environment in the last 20 years. A review finding by
(Pandey et al., 2017) also focused on the investigation of intelligent navigation methods of WMR that had been
developed. This finding included artificial potential field method, grids, Dijikstra algorithm, Voronoi graph,
vision sensors, fuzzy logic, neural network, genetic algorithm, and simulated annealing algorithm. The
method in (Song et al., 2023) was concerned with the path of MR in a dynamic environment by incorporating
the improved ant colony optimization ACO and dynamic window method DWA algorithm. Moreover, an
optimization algorithm was used by (Chai et al., 2024) to achieve online tuning of kinematic and dynamic
controllers, ensuring fast and accurate convergence of the trajectory tracking error. Work of (Ajeil et al., 2020)
deals with the design of intelligent path planning algorithms for a mobile robot in static and dynamic
environments based on swarm intelligence optimization. A modification based on the age of the ant is
introduced to standard ant colony optimization, called aging-based ant colony optimization (ABACO). Paper
(Sun et al., 2020) proposes a robust nonsingular terminal sliding mode (NTSM) control scheme for the path-
following problem of an MWOMR. First, a plant model is identified as a second-order state-space equation
with four inputs and three outputs to describe the MWOMR’s path-tracking kinematics and dynamics.
Afterwards, a multi–input–multi–output NTSM controller is designed for the MWOMR, and the stability of
the NTSM control system is verified by means of Lyapunov function. (Yuan et al., 2019) proposed an extended
state observer‐based sliding mode control (ESO‐SMC) strategy for trajectory tracking of a four mecanum
wheeled mobile platform (FMWMP) with unknown disturbances and model uncertainties (UDMU)
considered. Especially, the extended state observer (ESO) is designed to estimate not only the UDMU but also
the unmeasured velocities of FMWMP. Based on the designed ESO, a sliding mode control (SMC) scheme is
utilised to ensure the tracking performance as expected. The target of (Tuan et al., 2024) research is to develop
a control method for mobile robots operating in dynamic and unknown environments with both static and
moving obstacles. The method utilizes a fuzzy logic approach based on a Lidar sensor to gather information
about the environment and make decisions for navigation. Approach (Sun et al., 2023) proposes a trajectory-
 84
International Journal of Technology 16(1) 82-96 (2025)

planning scheme and a trajectory-tracking control strategy for a Mecanum-wheeled omnidirectional mobile
robot by using artificial potential field and discrete integral terminal sliding mode, respectively. According to
the positions of the robot, target, and obstacles, a reference an obstacle-avoidance trajectory is planned and
updated iteratively by utilizing artificial potential field functions. Afterward, a discrete integral terminal
sliding mode control algorithm is designed for the omnidirectional mobile robot such that the robot can track
the planned trajectory accurately (Husain et al., 2024) proposed and designed a strong control scheme for
vehicle steer by wire using an integral sliding mode control based on barrier function. This control scheme
improved the strength of the controllers better than that of conventional SMC or integral SMC. For the
problem of mobile robot’s path planning under the known environment, a path planning method of mixed
artificial potential field (APF) and ant colony optimization (ACO) based on grid map has been proposed by
(Chen and Liu 2019). (Bernardo et al., 2023) implemented hierarchical control structure using behavior tree
(BT) to improve the flexibility and adaptability of an omni-directional MR. This method improved point
stabilization and guided the robot to any configuration in the workspace while satisfying state constraints
(obstacle avoidance) as well as input constraints (motor limits). A novel adaptive fuzzy prescribed
performance control strategy is proposed in (Ding et al., 2024) to deal with the problem of fast accurate
trajectory tracking. Instead of the approach angle or the azimuth angle, a pair of heading- and orientation-
related auxiliary variables is adopted to deal with the underactuation of the WMR. In place of the transverse
function, the barrier function is combined with the adaptive fuzzy logic system to cope with the unknown
WMR dynamics.
(Tilahun et al., 2023) proposed trajectory tracking control system using a dual-loop method. In this method,
inner loop controlled the dynamics through Adaptive Fuzzy Sliding Mode Control (AFSMC), while outer
loop handled kinematics using Adaptive Neuro-Fuzzy Inference System (ANFIS). Generally, the result
showed that AFSMC with ANFIS was superior in trajectory tracking for the examined system compared to
other controllers. (Yan et al., 2024) designed a local tracking controller that ensured the linear velocity input
remained positive, and confirmed the position tracking error was convergent to zero. (Azar et al., 2023)
comprehensively examined and analyzed a proposed sliding mode higher–order extended state observer
(SMHOESO) convergence using Lyapunov method. According to this analysis, SMHOESO was
asymptotically stable, and estimation error was significantly reduced under real-world conditions. The
method proposed by (Abukhalil et al., 2020) used pattern recognition of two laser pointers in the view of a
single camera to detect objects around the robot. Additionally, distance and angle to the objects were
measured using Lagrange interpolation formula applied separately to each laser projection in the framed
image. By reviewing the methods used in previous research on movable obstacle avoidance, including those
discussed here, this topic still presents many challenges that require further investigation. There is a
significant research gap in the field of movable obstacle avoidance, providing more opportunities for further
exploration and experiments in this area. Therefore, this research is considered a step in the direction of
innovation in the field.

2. Mobile Robot Kinematic and Dynamic Modeling

In this section, kinematic and dynamic model of WMR were obtained. The movement of the
proposed WMR (Figure 1) was only in X0Y0 plane. Following the discussion, two different
coordinate frames were evident during the research. The first was local coordinate frame (OXY) of
the robot where X-axis represented direction of forward movement and Y-axis signified lateral
movement as shown in Figure 1. This frame was dynamic and changed as the robot moved, which
was related to the second or global coordinate frame (oX0Y0). The velocity of WMR in its local
coordinate frame consisted of linear velocity in X and Y direction and a rotation around the center.
This velocity was written as, 𝑉𝑟 = (𝑋̇ 𝑌̇ 𝜃̇ )𝑇 .
 85
International Journal of Technology 16(1) 82-96 (2025)

Figure 1 WMR configuration

Due to the rotation between WMR local and global coordinate frame with a rotation angle θ and
rotation matrix 𝐺𝐿𝑅, velocity 𝑉𝑟 was transferred to global coordinate frame as 𝑉𝑟𝑔 = (𝑋̇𝑜 𝑌𝑜̇ 𝜃̇ )𝑇 . In
addition, the transformation of the velocity was written as Equation 1 (Alwan et al., 2024).

𝑉𝑟𝑔 = 𝐺𝐿𝑅 𝑉𝑟 (1)

MR velocity expressed in local coordinates system was used to develop kinematic and dynamic
mathematical model concerning angular velocity as well as torques of the wheels. The velocity
acting on each helm of the mobile wheels consisted of four components including linear velocity (
𝑋̇ , 𝑌̇) due to WMR movement. Other components included tangential velocity (𝜔 𝑟) caused by
angular velocity of the wheel, roller tangential velocity (𝜔 𝑟) from angular velocity of the roller
which was perpendicular to roller axis (45⁰) with the robot X-axis. Finally, linear velocity was caused
by the rotation of WMR around its center point. WMR inverse kinematic model was obtained by
summing up this velocity and was written as Equation 2(Alwan et al., 2024):

(𝜔) = 𝐽𝑉𝑟 (2)

1 −1 −(𝑎 + 𝑏)
1 11 (𝑎 + 𝑏)
Where J was the inverse Jacobian matrix; 𝐽 = [ ], and 𝝎 represented the
1 1
𝑟𝑤𝑖−(𝑎 + 𝑏)
1 −1 (𝑎 + 𝑏)
wheel angular velocity matrix 𝜔 = (𝜔1 𝜔2 𝜔3 𝜔4 )𝑇 .

The forward kinematics equation in the robot local coordinates system was written as Equation 3
(Alwan, 2020a),

𝑉𝑟 = 𝐽+ (𝜔) (3)
The forward kinematics equation for WMR in global coordinate system was expressed in Equation
4.
 86
International Journal of Technology 16(1) 82-96 (2025)

𝑉𝑟𝑔 = 𝐺𝐿𝑅 𝐽+ (𝜔) (4)

Dynamic modeling of WMR was obtained using Lagrange formulation. Significantly, the velocity
of mass center of WMR in the robot local coordinate system was evaluated according to the theory
of (Alwan, 2020b) as written in Equation 5.

𝑐𝑜𝑠𝜃 𝑠𝑖𝑛𝜃 𝑋̇𝑜

𝑣𝐶 = ( )[ ] (5)
−𝑠𝑖𝑛𝜃 𝑐𝑜𝑠𝜃 𝑌𝑜̇

WMR moved in a horizontal plane and not in z-axis direction, leading to no potential energy.
Kinetic energy of the robot mechanism was caused by linear and rotational velocity of MR as well
as four Mecanum wheels(𝑣𝑐 , 𝜃̇ , 𝑟𝜔, 𝜔). Moreover, kinetic energy equation was obtained as follows
(Equation 6) (Alwan, 2020b).
1 2
𝑇 = 2 [𝑚𝑝 𝑣𝐶𝑇 𝑣𝐶 + 𝐼𝑝 (𝜃̇ ) + 4[𝑚𝑤 (𝑟𝜔)2 + 𝐼𝑤 𝜔2 ]] (6)

Where (mp , mw , Ip and Iw ) were the robot platform mass, wheel mass, robot platform moment of
inertia around its mass center, and wheel moment of inertia, respectively.
The dynamic of WMR presented the derivation of Lagrange equation, which was written as
Equation 7.
𝑑 𝜕𝑇 𝜕𝑇
( )− ( ) = 𝐹𝑖 i=1,2,3,4 (7)
𝑑𝑡 𝜕𝑞̇ 𝑖 𝜕𝑞𝑖
Where 𝑞𝑖 , 𝑞̇ 𝑖 represented the inputs or angular displacement and angular velocity of i wheel.
WMR general form dynamic equation was expressed as Equation 8 (Hasan and Alwan, 2021).
1
𝑀(𝑞)𝑞̈ + 𝐶(𝑞, 𝑞̇ )𝑞̇ + 𝐵𝑇 𝑆 𝑓 = 𝐵𝑇 𝜏 (8)
𝑟

Where 𝑞̇ = (𝑋̇𝑜 𝑌0̇ 𝜃̇ )𝑇 , 𝑞̈ = (𝑋̈𝑜 𝑌0̈ 𝜃̈ )𝑇 ,M was the mass and inertia matrix, C signified matrix
of coriolis as well as centripetal effect, and B was the input transformation matrix. The elements of
each matrix were represented as Equations 9, 10, and 11 (Alwan et al., 2024).
0 0 𝑚𝑟 𝜃̇ (ℎ1 𝑐𝑜𝑠𝜃 + ℎ2 𝑠𝑖𝑛𝜃)
𝐶 = [0 0 𝑚𝑟 𝜃̇ (ℎ1 𝑠𝑖𝑛𝜃 + ℎ2 𝑐𝑜𝑠𝜃)] (9)
0 0 0
(𝑐𝑜𝑠𝜃 + 𝑠𝑖𝑛𝜃) (𝑠𝑖𝑛𝜃 − 𝑐𝑜𝑠𝜃) −(𝑎 + 𝑏)
(𝑐𝑜𝑠𝜃 − 𝑠𝑖𝑛𝜃) (𝑠𝑖𝑛𝜃 + 𝑐𝑜𝑠𝜃) (𝑎 + 𝑏)
𝐵=[ ] (10)
(𝑐𝑜𝑠𝜃 − 𝑠𝑖𝑛𝜃) (𝑠𝑖𝑛𝜃 + 𝑐𝑜𝑠𝜃) −(𝑎 + 𝑏)
(𝑐𝑜𝑠𝜃 + 𝑠𝑖𝑛𝜃) (𝑠𝑖𝑛𝜃 − 𝑐𝑜𝑠𝜃) (𝑎 + 𝑏)
𝑠𝑔𝑛(𝜑̇ 1 ) 0 0 0
0 𝑠𝑔𝑛(𝜑̇ 2 ) 0 0
𝑆=[ ] (11)
0 0 𝑠𝑔𝑛(𝜑̇ 3 ) 0
0 0 0 𝑠𝑔𝑛(𝜑̇ 4 )

3. Motion Control Design for Trajectory Tracking

Backstepping method was adopted to execute the controller during the analysis. The state vector
of the generalized coordinate was written as follows (Equation 12).
𝑇
𝑞 = [𝑞𝑎𝑇 𝑞̇ 𝑎𝑇 ]𝑇 = (𝑋𝑜 𝑌𝑜 𝜃𝑜 𝑋̇𝑜 𝑌𝑜̇ 𝜃̇𝑜 ) (12)
 87
International Journal of Technology 16(1) 82-96 (2025)

Based on backstepping controller 𝑞̇ 𝑎 was a subsystem which was equal to virtual input (u1).
Trajectory tracking error equation of WMR movement represented the error between desired and
actual trajectories as written in Equation 13.

𝑋0𝑑 − 𝑋0𝑎𝑐
𝑒1 = 𝑞𝑑 − 𝑞𝑎𝑐 = (𝑒𝑥 𝑒𝑦 𝑒𝜃 )𝑇 = ( 𝑌0𝑑 − 𝑌0𝑎𝑐 ) (13)
𝜃𝑑 − 𝜃𝑎𝑐

𝑇 𝑇
Where 𝑞𝑑 = [𝑋0𝑑 𝑌0𝑑 𝜃𝑑 ] was desired robot trajectory and 𝑞𝑎𝑐 = [𝑋0𝑎𝑐 𝑌0𝑎𝑐 𝜃𝑎𝑐 ] represented
actual robot trajectory. The derivative of trajectory tracking error equation (13) led to Equation 14.
𝑒̇1 = 𝑞̇ 𝑎 − 𝑞̇ 𝑑 = 𝑢1 − 𝑞̇ 𝑑 (14)
For WMR stability testing, Lyapunov expression was considered as Equation 15:

1
𝑉1 = 2 𝑒1𝑇 𝑘1 𝑒1 (15)

Where k 1  R 33 was symmetrical and positive. During this analysis, Equation 15 was
differentiated concerning time as 16.
1 1 1 1
𝑉1̇ = 2 𝑒̇1𝑇 𝑘1 𝑒1 + 2 𝑒1𝑇 𝑘1 𝑒̇1 = 2 𝑒1𝑇 𝑘1 𝑒̇1 = 2 𝑒1𝑇 𝑘1 (𝑢1 − 𝑞̇ 𝑑 ) (16)
For stable system, Equation 17 was expressed as,
𝑢1 = 𝑞̇ 𝑑 − 𝑒1 (17)
Equation 18 was written as,
1
𝑉1̇ = − 2 𝑒1𝑇 𝑘1 𝑒̇1 ≤ 0 (18)

Equation 18 showed that the system was asymptotically stable.

4. Obstacle Avoidance Process

MR obstacle avoidance decisions were typically based on real-time data concerning the position
and velocity of obstacle. This critical information was obtained through sensors during the detection
process. In this research, six HC-SR04 ultrasonic sound sensors were used to detect obstacle position
and velocity. Two sensors were placed in the front direction of the robot, while the remaining ones
were positioned at angles of +45⁰, +90⁰, -45⁰, and -90⁰. The sensors which were placed in the front
direction determined how the obstacle moved. Figure 2 showed the locations of the devices on the
robot body and field of view. Moreover, the field of view of these sensors was 15⁰, and the sensing
range was from 2 cm to 400 cm.
The obstacle avoidance system of the robot focused on detecting and evading obstructions in a
circle less than 0.5 meters radius around the robot center point. In this research, the obstacle
avoidance algorithm showed that all obstacle were circular in shape and moved at a constant
velocity perpendicular to the path of the robot.
 88
International Journal of Technology 16(1) 82-96 (2025)

Figure 2 The field of view and sensors locations of the robot

The Major points in the obstacle avoidance algorithm were as follows:

• The robot often moved forward following the desired path since there were no obstacles.
• When an obstacle was detected on front side of the robot, position, and velocity of the obstruction
were determined using the sensors located in that area.
• The optimal alternative path in which the robot moved was determined based on the direction
of how obstacle moved and the signals received from other sensors to identify obstacle-free path.
• The robot avoided obstacle by changing its original desired path to feasible alternative route,
which was possible due to the use of Mecanum wheels.
• The feasible alternative paths of the robot were arranged according to proposed priority. For
example, when the detected obstacle was coming from right to left, the first priority of the robot
was to continue its movement until the obstacle passed. Once collision was unavoidable, the
robot followed the second priority, which was changing its desired route to feasible alternative
path at direction of an angle of -45⁰. However, when collision could not be avoided, the robot
followed the third priority, which was changing its desired path to an alternative route at
direction of an angle of -90⁰. The robot sequentially moved to its fourth and fifth priority and then
stopped completely at the final designated position, waiting for obstacle to pass.
• The feasible alternative path directions in which the robot moved to avoid collision were
arranged in the following directions including +45⁰, +90⁰, -45⁰, and -90⁰. When none of these paths
were suitable, the robot stopped moving. Table 1 showed the priority of the feasible alternative
path directions.
• Figure 3 also showed the alternative WMR direction priority in different cases of obstacle
avoidance.
• After avoiding obstacle, the robot returned to tracking the desired path.
• Obstacle avoidance algorithm methodology was shown in Figure 4
 89
International Journal of Technology 16(1) 82-96 (2025)

Table 1 Alternative robot paths directions priority in obstacle avoidance process

Obstacle Alternative Robot Path Directions
Direction +90° +45° 0° -45° -90° stop
From right priority priority priority priority priority priority
to left 4 5 1 2 3 6
From left to priority priority priority priority priority priority
right 3 2 1 5 4 6

Figure 3 The alternative WMR directions priority in different cases of obstacle avoidance

Figure 4 The obstacle avoidance algorithm methodology

For detecting a movable obstacle and determining its velocity as well as direction, WMR needed
to sense the obstruction at two different positions in a known time interval (Figure 5). At each point
obstacle center position (CP) in global coordinate frame was determined based on WMR CP and the
distance between obstacle and the robot measured by the sensor. When the obstacle positions at the
 90
International Journal of Technology 16(1) 82-96 (2025)

two points were known, the period required to move the obstruction from the first to second point,
its velocity and direction, was evaluated as Equation 19.
2
√([xob (i+1)−xob (i))2 +(yob (i+1)−yob (i)])2
Vob = ∆t
(19)
∆𝑦𝑜𝑏
𝜃𝑜𝑏 = 𝑎𝑟𝑡𝑎𝑛 ( )
∆𝑥𝑜𝑏
The distance between MR and obstacle was more than 500 mm to avoid collision. Relating to the
process, this distance was obtained as follows (Equation 20).

𝑑 = √[𝑋0 (𝑡) − 𝑋𝑜𝑏 (𝑡)]2 + [𝑌0 (𝑡) − 𝑌𝑜𝑏 (𝑡)]2 (20)

Figure 5 Detection of movable obstacle

5. Motion Control and Simulation Results

The aim of the control method was finding an optimal trajectory for MR considering the three
main criteria, including short trajectory, smooth trajectory and obstacle–free path, respectively.
Moreover, the cost functions used to represent short trajectory 𝒇𝟏 as well as its smoothness 𝒇𝟐 cost
functions was written as Equations 21 and 22 (Hasan and Alwan, 2022).
f1 (X0 , Y0 ) = ∑N−1 2
i=1 √[X 0 (t + 1) − X 0 (t)] + [Y0 (t + 1) − Y0 (t)]
2 (21)

Y(t+1)−Y(t)
f2 (X0 , Y0 ) = ∑N−1
i=1 arctan X (22)
0 (t+1)−X0 (t)

During the analysis, Artificial Bee Colony (ABC) optimization algorithm was adopted. By
initializing the parameters as well as proposing a random number of populations (SN), the process
was repeated to reach the maximum cycle number, and the solution was written as Equation 23
(Hasan and Alwan, 2022).
𝑥𝑖𝑗 = 𝑥𝑚𝑖𝑛𝑗 + 𝑟𝑎𝑛𝑑 [0,1](𝑥𝑚𝑎𝑥𝑗 − 𝑥𝑚𝑖𝑛 ) (23)
Where і= 1,2..SN, j= 1,2..D, D, represented the number of the optimized parameters, rand (0,1) is a
real number in interval [0,1], 𝑥𝑚𝑎𝑥 represented upper bounds of 𝑥𝑖𝑗 , and 𝑥𝑚𝑖𝑛 was lower bounds of
𝑥𝑖𝑗 . Following this discussion, a new solution was evaluated as Equation 24 (Hasan and Alwan,
2022):
𝑣𝑖 = 𝑥𝑖𝑗 + 𝜑𝑖𝑗 (𝑥𝑖𝑗 − 𝑥𝑖𝑘 ) (24)
Where k ϵ [1, 2…SN]. When the nectar amount of (𝑣𝑖𝑗 ) was better than (𝑥𝑖𝑗 ), (𝑥𝑖𝑗 ) was changed by
(𝑣𝑖𝑗 ) and (𝑣𝑖𝑗 ) became the new population number.
 91
International Journal of Technology 16(1) 82-96 (2025)

The controller algorithm generated suitable improvements to control MR wheel angular velocity
and the robot movement along original or alternative trajectories. During the algorithm testing,
simulations were performed to show the motion of the robot along a question-sign-shaped
trajectory. Moreover, simulation results of equation (17) were performed in MATLAB Simulink
program. The following geometric dimensions were adopted to perform numerical simulation
movement of the robot including Mecanum wheel angle δ = π/4[rad], WMR represented as a
cylinder of R = 0.25[m] and the obstacle was as a colander of a radius Rob = 0.15[m]. Figure 6 showed
WMR control architecture diagram used to simulate the results of motion control and obstacle
avoidance.

Figure 6 Control architecture diagram

Simulation and validation of two examples of obstacle avoidance were presented in this section.
Example–1: Avoidance of movable obstacle was implemented in this example, where WMR
operated in unknown environment containing two movable obstacle, and each of the obstacle
intersected the desired path in two points. When one or more of the sensors detected an obstacle,
signal was sent to the controller to determine the position of obstacle center. After a period, this
procedure was repeated to determine the position of the center. Based on the two records, obstacle
velocity and angle of its direction were determined. When the distance between WMR and obstacle
was detected to be less than 500 mm, MATLAB initiated avoidance algorithm. This algorithm sent
the information to the controller, adjusting MR wheels angular velocity and enabling it to follow an
alternative path. The selected path was determined based on the obstacle velocity, movement
direction, and priority of feasible alternative paths. Results of this analysis showed that the path
length travelled by WMR without obstacle was 1026.7 cm in 37.17 seconds, while length travelled
with movable obstacle was 1162.5 cm in 38 seconds, respectively. Table 2 showed the added path
length and time as a result of avoiding each of the four obstacle. Additionally, trajectory tracing
error between WMR desired and its actual path to avoid the collision was shown in Figure 7.
 92
International Journal of Technology 16(1) 82-96 (2025)

Simulated result of movable obstacle avoidance process which included WMR desired and actual
path to avoid the collision, as well as movable obstacle paths was shown in Figure 8. The results of
this analysis showed that the percentage of maximum additional time to avoid movable obstacle
was 0.7% of the total time. In addition, the percentage of maximum additional length to avoid this
obstacle was 3.58% of the total desired length.

Table 2 Path length and time increase from movable obstacle avoidance
Obstacle Distance increase Percentage of Time Percentage of
number [m] distance increase increase [s] time increase
1 0.368 3.58% 0.26 0.7%
2 0.146 1.42% 0.103 0.28%
3 0.337 3.28% 0.239 0.64%
4 0.279 2.72% 0.197 0.53%

Figure 7 Trajectory tracking error and avoidance of movable obstacle

Figure 8 Simulated results of movable obstacle avoidance trajectory

Example–2: This example was allocated to static obstacle avoidance during the analysis. WMR
worked in an unknown environment containing four static obstacle, located in various positions.
When one or more of the sensors detected an obstacle, signal was sent to the controller in
determining the position of obstacle center. Moreover, when the detected distance between WMR
and obstacle was less than 500 mm, MATLAB initiated the avoidance algorithm and sent
information to the controller, adjusting MR wheels angular velocity to follow an alternative path
based on obstacle position and priority of feasible alternative paths. The results of this process
showed that path length travelled by WMR without obstacle was 1098cm in 38.88 seconds, while
with static obstacle was 1244.5 cm in 40 seconds, respectively. Table 3 showed the added path length
and time as a result of avoiding each of the four obstacles. Trajectory tracing error between WMR
 93
International Journal of Technology 16(1) 82-96 (2025)

desired and its actual path to avoid the collision was shown in Figure 9. Additionally, simulated
result of static obstacle avoidance process which included WMR desired and actual WMR path to
avoid the collision, as well as fixed obstacle poisons was shown in Figure (10). Results of the analysis
showed that the percentage of maximum additional time to avoid fixed obstacle was 1.15% of the
total time. In addition, the percentage of maximum additional length to avoid this obstacle was
5.52% of the total desired length. In comparison with previous research under the same conditions,
this investigation found that in (Hasan and Alwan, 2022) and (Hasana and Alwan, 2021), MR
traveled a path of 516.15cm in 25.989 seconds, and 1451.06cm in 73.5 seconds, respectively.
Meanwhile, MR traveled 1098cm in 38.88 seconds in this research, signifying that WMR avoided the
obstacle faster than in other methods.

Table 3 Path length and time increase from static obstacle avoidance
Obstacle Distance increase Percentage of Time Percentage of
number [m] distance increase increase [s] time increase
1 0.189 1.72% 0.139 .036%
2 0.203 1.85% 0.149 .038%
3 0.505 4.6% 0.371 .095%
4 0.626 5.52% 0.46 1.18 %

Figure 9 Trajectory tracking error while avoidance of static obstacle

Figure 10 Simulated results of static obstacle avoidance trajectory

 94
International Journal of Technology 16(1) 82-96 (2025)

The analysis signified that the tracking error in the previous two examples increased when
avoiding obstacle. This process was normal, but a relatively large error occurred when crossing 90⁰
angles in the track. The error was the result of basic error represented by displacement from the
intended aimed point, which was approximately 0.36m, and the fault was rectified at the end of
track. Following the discussion, displacement occurred because the aimed point kept changing at
every moment, but response movement of the robot was slower than the speed of change, leading
to displacement.

6. Conclusions
In conclusion, this research developed MR equipped with four Mecanum wheels and formulated
its kinematic as well as dynamic models. Trajectory tracking control system was also designed using
obstacle avoidance algorithm for both stationary and movable obstacle. The performance of the
robot was thoroughly evaluated, and results showed that the machine successfully achieved its
main objectives of avoiding all encountered stationary or movable obstacle. Additionally, the
outcome showed that the avoidance of the four static obstacle increased path length by 13% and
2.87% for the time to reach target. As for the four movable obstacle, path length increased by 11%
and 2.15% for the time to reach target. Trajectory tracking error during the analysis was significant
only during obstacle avoidance maneuvers, showing the effectiveness of the control system in
normal conditions. However, a lateral displacement error of 0.36 meters was observed, attributed
to slow response of the robot to dynamically changing target points. The error occurred only at
inflection points because the running of obstacle avoidance algorithm slowed down the response at
these points. This outcome signified that, although the control and avoidance algorithms were
strong, further optimization would be needed to improve responsiveness of the robot to rapidly
changing target points. Future research should focus on improving the responsiveness and accuracy
of trajectory tracking system to mitigate lateral displacement errors. The investigations should also
focus on achieving more precise navigation and testing the improved methods with MR as well a
movable obstacle under more complex conditions.

Acknowledgements
The authors are grateful to everyone who provided support from University of Technology as well as
High School of Automation and Robotics at Peter the Great Saint Petersburg Polytechnic University.

Nomenclature
WMR Wheeled Mobile Robot
FMWMR Four Mecanum Wheeled Mobile Robot
LQR Linear Quadratic Regulator
ACO Ant Colony Optimization
DWA Dynamic Window Approach
ABC Artificial Bee Colony
PID Proportional Integral Derivative
ESO Extended State Observer
SMHOESO Sliding Mode Higher – Order Extended State Observer
SMC Sliding Mode Control
MPC Model Predictive Control

Author Contributions
Hassan M. Alwan; Conceptualization, Methodology, Derivation of Robot Kinematic and Dynamic Models
and Formal Analysis, Volkov A.N.; Review and Supervision, Ali Shbani and Kochneva O.V.; Results
Simulation, Software and Project Administration. All authors have read and agreed to the published version
of the manuscript.
 95
International Journal of Technology 16(1) 82-96 (2025)

Conflict of Interest
The authors confirm that there is no conflict of interest to declare for this publication

References
Abdelwahab, M, Parque, V, Elbab, AMF, Abouelsoud, A & Sugano, S 2020, ‘Trajectory tracking of wheeled
mobile robots using z-number based fuzzy logic’, IEEE Access, vol. 8, pp. 18426-18441,
[Link]
Abed, MS, Lutfy, OF & Al-Doori, QF 2021, ‘A review on path planning algorithms for mobile robots’,
Engineering and Technology Journal, vol. 39, no. 5, pp. 804-820, [Link]
Abukhalil, T, Alksasbeh, M, Alqaralleh, B & Abukaraki, A 2020, ‘Robot navigation system using laser and
monocular camera’, Journal of Theoretical and Applied Information Technology, vol. 98, pp. 714-724
Ajeil, FH, Ibraheem, IK & Azar, AT 2020, ‘Grid-based mobile robot path planning using aging-based ant
colony optimization algorithm in static and dynamic environments’, Sensors, vol. 20, no. 7, article 1880,
[Link]
Al-Mallah, M, Ali, M & Al-Khawaldeh, M 2022, ‘Obstacle avoidance for mobile robot using type-2 fuzzy
logic controller’, Robotics, vol. 11, no. 6, article 130, [Link]
Alwan, HM 2020a, ‘Dynamic analysis modeling of a holonomic wheeled mobile robot with mecanum
wheels using virtual work method’, Journal of Mechanical Engineering Research and Developments , vol. 43,
no. 6, pp. 373-380
Alwan, HM 2020b, ‘Path tracking simulation of a wheeled mobile robot with three mecanum wheels’,
International Review of Mechanical Engineering, vol.14, no. 8, pp. 516-522,
[Link]
Alwan, HM, Nikolaevic, VA, Hasan, SF & Vladmerovna, KO 2024, ‘Kinematic and dynamic modeling
based on trajectory tracking control of mobile robot with mecanum wheels’, International Journal of
Technology, vol. 15, no. 5, pp. 1473-1486, [Link]
Azar, AT, Abed, AM, Abdul-Majeed, FA, Hameed, IA, Jawad, AJM, Abdul-Adheem, WR, Ibraheem, IK &
Kamal, NA 2023, ‘Design and stability analysis of sliding mode controller for non-holonomic differential drive
mobile robots’, Machines, vol. 11, no. 4, article 470, [Link]
Bernardo, R, Sousa, JMC, Botto, MA & Goncalves, PJS 2023, ‘A novel control architecture based on behavior
trees for an omni-directional mobile robot’, Robotics, vol. 12, no. 6, article 170,
[Link]
Chai, B, Zhang, K, Tan, M & Wang, J 2024, ‘An optimal robust trajectory tracking control strategy for the
wheeled mobile robot’, International Journal of Control, Automation and Systems, vol. 22, pp. 1050-1065,
[Link]
Chen, G & Liu, J 2019, ‘Mobile robot path planning using ant colony algorithm and improved potential
field method’, Computational Intelligence and Neuroscience Journal, vol. 2019, article 1932812,
[Link]
Dang, TV & Bui, NT 2023, ‘Multi-scale fully convolution network-based semantic segmentation for mobile
robot navigation’, Electronics, vol. 12, no. 3, article 533, [Link]
Dang, T-V & Bui, N-T 2023, ‘Obstacle avoidance strategy for mobile robot based on monocular camera’,
Electronics Journal, vol. 12, no. 8, article 1932, [Link]
Ding, W, Zhang, JX & Shi, P 2024, ‘Adaptive fuzzy control of wheeled mobile robots with prescribed
trajectory tracking performance’, IEEE Transactions on Fuzzy Systems, vol. 32, no. 8,
[Link]
Feng, S, Sebastian, B & Ben-Tzvi, P 2021, ‘A collision avoidance method based on deep reinforcement
learning’, Robotics, vol. 10, no. 2, article 73, [Link]
Hasan, SF & Alwan, HM 2022, ‘Local path planning of a four mecanum wheeled mobile robot based on
new modified ultrasonic sensors with experimental implementation’, International Journal of Mechanical
Engineering, vol. 7, no. 1, pp. 4621-4627
Hasan, SF& Alwan, HM 2021, ‘Obstacle avoidance of wheeled mobile robot by using modified artificial bee
colony optimization’, Design Engineering, vol. 7, pp. 3713-3727
Hasana, SF & Alwan, HM 2021, ‘Modelling and control of wheeled mobile robot with four mecanum
wheels’, Engineering and Technology Journal, vol. 39, no. 5, pp. 779-789
 96
International Journal of Technology 16(1) 82-96 (2025)

Husain, SS, Al-Dujaili, AQ, Jaber, AA, Humaidi, AJ & Al-Azzawi, RS, 2024, ‘Design of a Robust Controller
Based on Barrier Function for Vehicle Steer-by-Wire Systems, ‘World Electric Vehicle Journal, 15, 17.
[Link]
Kuo, P-L, Wang, C-H, Chou, H-J & Liu, JS 2018, ‘A real-time hydrodynamic-based obstacle avoidance
system for non-holonomic mobile robots with curvature constraints’, Applied Sciences Journal, vol. 8, no. 11,
article 2144, [Link]
Pandey, A, Pandey, S & Parhi Dr 2017, ‘Mobile robot navigation and obstacle avoidance techniques: a
review’, International Robotics &Automation Journal, vol. 2, no. 3, pp. 96-105
Rusdinar, A, Purnama, I, Fuadi, AZ, Adiluhung, H, Wicaksono, M, Risnanda & Ningrum, RA 2021,
‘Automated ultraviolet c light mobile robot for room sterilization and disinfection’, International Journal of
Technology, vol. 12, no. 4, pp. 854-864, [Link]
Song, B, Tang, S & Li, Y 2023, ‘A new planning integrating improved aco and dwa algorithms for mobile
robots in dynamic environments’, Mathematical and Biosciences and Engineering Journal , vol. 21, no. 2, pp.
2189-2211, [Link]
Sun, Z, Hu, S, Miao, X, Chen, B, Zheng, J, Man, Z & Wang, T 2023, ‘Obstacle - avoidance trajectory planning
and sliding mode-based tracking control of an omnidirectional mobile robot’, Frontiers In Control
Engineering, vol. 4, article 1135258, [Link]
Sun, Z, Xie, H, Zheng, J, Man, Z, & He, D 2020, ‘Path-following control of mecanum-wheels omnidirectional
mobile robots using nonsingular terminal sliding mode’, Mechanical Systems and Signal Processing, vol. 2,
no. 2, article 107128, [Link]
Tilahun, AA, Desta, TW, Salau, AO & Negash, L 2023, ‘Design of an adaptive fuzzy sliding mode control
with neuro-fuzzy system for control of a differential drive wheeled mobile robot’, Cogent Engineering, article
2276517, [Link]
Tuan, PM, Tai, ND, Huy TQ & Thinh, NT 2024, ‘Flexible path planning of mobile robot for avoiding the
dynamic obstacle using fuzzy controllers’, International Journal of Mechanical Engineering and Robotics
Research, vol. 13, no. 1, pp 126-132, doi: 10.18178/ijmerr.13.1.126-132
Wang, H, Ma, X & Zhu, L 2022, ‘Obstacle avoidance path planning of mobile robot based on improved
DWA’, Journal of Physics: Conference Series, article 2383, doi:10.1088/1742-6596/2383/1/012098
Wang, W, Zhao, J, Li, Z & Huang, J 2021b, ‘Smooth path planning of mobile robot based on improved ant
colony algorithm, Journal of Robotics, vol. 202, [Link]
Wang, Y, Li, X, Zhang, J, Li, S, Xu, Z & Zhou, X 2021a, ‘Review of wheeled mobile robot collision avoidance
under unknown environment’, Science Progress Journal, vol. 104, no. 3,
[Link]
Yan, L, Ma, B & Jia, Y 2024, ‘Trajectory tracking control of nonholonomic wheeled mobile robots using only
measurements for position and velocity’, Automatica, vol. 159, article 111374,
[Link]
Yuan, Z, Tian, Y, Yin, Y, Wang, S, Liu, J & Wu, L 2019, ‘Trajectory tracking control of a four mecanum
wheeled mobile platform: an extended state observer-based sliding mode approach’, The institution of
Engineering and Technology, vol. 4, no. 5, pp. 1-12, [Link]
Zghair, NAK & Al-Araji, AS 2023, ‘Dynamic obstacle avoidance algorithm for autonomous mobile robot’,
Iraqi Journal of Computers, Communications, Control and System Engineering, vol.23, no. 2, pp. 63-82