Adaptive brake force distribution (BFD) is critical for the safety and efficiency of electric commercial vehicles (ECVs), yet the integration of thermal constraints, real-time control, and stability requirements remains fragmented. Following PRISMA 2020, a systematic search in IEEE Xplore and Web of Science (2016–2026) was performed, yielding 156 core studies classified into five thematic groups: fundamentals, MPC-based control, PSO, integrated frameworks, and validation methods. The analysis reveals a tripartite fragmentation: 27 MPC studies neglect thermal constraints; 22 PSO studies lack real-time adaptability; 52 studies target commercial vehicles, but none validate front–rear redistribution under concurrent thermal and stability constraints. Only two studies address thermal-aware distribution, and seven consider pneumatic delay compensation. Publication volume accelerates, indicating rising interest. The literature treats thermal management, real-time redistribution, and stability-constrained control as separate problems, leaving a critical gap: online correction of the ideal distribution curve under coupled thermal-stability constraints for ECVs is absent. To fill this gap, a hierarchical MPC-PSO framework is proposed, where PSO (1–5 Hz) adaptively tunes MPC weights based on vehicle state, and MPC (20–50 Hz) executes optimal control integrating regenerative, magnetic, and eddy current braking. This research aims to achieve two main objectives: to propose a new integrated framework that addresses the overlooked interrelationship between the stability constraints of ECVs and thermal constraints and to systematically reveal the state of integrative constraints on adaptive brake force distribution in electric commercial vehicles, with a focus on real-time control and stability requirements.

The dynamic distribution of braking forces between front and rear axles in electric commercial vehicles represents a critical multi-objective optimization challenge requiring simultaneous satisfaction of regulatory safety compliance, regenerative energy recovery, thermal stability, and actuator coordination under varying load and road conditions. This paper addresses this challenge through the development and experimental validation of an integrated adaptive brake force distribution strategy combining model predictive control (MPC) with Particle Swarm Optimization (PSO) within a unified framework that ensures compliance with ECE Regulation No. 13. A comprehensive experimental test bench was designed and instrumented, integrating three independent braking mechanisms: magnetic brakes with front and rear torque coefficients of 4.73 N·m/A and 3.65 N·m/A, respectively; an eddy current retarder with coefficient $k_0=$ 2.220 × 10⁻⁴ N·m·s/(A²·rad), producing braking torque that is quadratic in excitation current and linear in rotor speed; a regenerative braking system with 82–90% efficiency; and a switchable magnetic clutch for FWD/4WD operation. The MPC controller was formulated with a prediction horizon $N_p$ = 20, control horizon $N_c$ = 5, and sampling time $T_s$ = 20 ms. PSO was employed for systematic tuning of MPC weights using 30 particles over 50 iterations with cognitive and social coefficients $c_1$ = $c_2$ = 2.0 and linearly decreasing inertia from 0.8 to 0.4. A vehicle state estimation module using Kalman Filtering was developed for real-time estimation of vehicle mass (<3% error), road slope (<0.3% error), and road friction coefficient (<5% error). Experimental validation across eight comprehensive test scenarios demonstrates that the PSO-optimized MPC controller achieves 43% reduction in front RMSE (from 2.65 Nm to 1.52 Nm), 44% reduction in rear RMSE (from 0.78 Nm to 0.44 Nm), 100% ECE R13 compliance (improved from 67.5%), 57% settling time improvement (from 4.2 s to 1.8 s), 92% overshoot reduction (from 67% to 5%), and average recovered energy improvement from 3.51 kJ to 4.04 kJ. The proposed framework provides a comprehensive solution for next-generation electric commercial vehicle brake management systems.