Research and Implementation of an Improved Message Scheduling Mechanism Based on Adaptive Genetic Algorithm

CAN bus is widely implemented in distributed control systems, primarily for industrial automation communications. The message scheduling mechanism involves coordinating data transmission among multiple nodes while avoiding collisions and ensuring timely delivery of critical information. The CAN protocol employs a non-destructive arbitration mechanism based on message identifiers, where lower identifier values indicate higher priority. When two nodes transmit simultaneously, higher-priority messages continue transmission while lower-priority messages enter a backoff state. However, this mechanism may lead to message latency, degraded control performance, and reduced resource utilization efficiency. The adaptive genetic algorithm enhances traditional genetic algorithms by addressing premature convergence and low search efficiency in later evolutionary stages caused by the loss of high-quality gene segments under conventional genetic operators. This approach enables dynamic adjustment of CAN data frame transmission sequences rather than relying solely on fixed priorities. MATLAB simulations were conducted using standard CAN data frame samples. Results demonstrate that the proposed improved cosine-based AGA optimizes the CAN message scheduling mechanism, achieving a significant reduction in data frame transmission time.