Research on Solving Planning Models Using an Improved Adaptive Genetic Algorithm

This paper presents a study on the optimization of a linear programming model aimed at maximizing total revenue while adhering to certain constraints. Due to the large solution space and the complexity of the solving process, we developed an improved adaptive genetic algorithm (AGA). The AGA incorporates strategies such as a stochastic selection of year-based crossover and random point mutation, along with dynamic adjustments of crossover and mutation probabilities. Additionally, it utilizes multi-matrix chromosome encoding and a constraint-based method for repairing infeasible solutions. This approach significantly enhances both the efficiency and quality of the solutions. The model and algorithm were implemented using Python, and the optimal strategies for the years 2024 to 2030 were successfully derived.