CRC32

基于遗传算法的PID参数优化的MATLAB程序，采用的是二进制编码方式，通过传递函数的离散经过适应度计算 选择复制 交叉 变异操作得到最优解。(PID parameters based on genetic algorithm optimization of the MATLAB program, using the binary encoding, by the discrete transfer function adapted calculation select Copy crossover, mutation operations to obtain the optimal solution.)

将控制表中的控制量转变为实际的控制量 控制表中的输出尺度变换后的输出 (The control table of the control amount is converted to the actual control amount output after the output of the scale transformation table)

双门限法,双门限法的改进和推广,相关法的端点检测,方差法的语音端点检测,谱距离法的端点检测，谱熵在端点检测中的应用(Double threshold method, double threshold method improvement and promotion, endpoint detection correlation method, endpoint detection variance method, the endpoint detection spectral distance method, the Spectrum Entropy in the endpoint detection)

Polar coding and decoding program simulation

Dijkstra最短路径算法，利用matlab实现，根据距离矩阵求解(Dijkstra shortest path algorithm)

精通matlab最优化计算随书带的源码，包括线性规划，整数规划，pso，GA等(Proficient in matlab optimization calculation with the source code with the book, including linear programming, integer programming, pso, GA etc.)

Single-Phase half-controlled bridge rectifier ALPHA=120

概率神经网络的分类预测-基于PNN变压器故障诊断(Classification and Prediction of Probabilistic Neural Networks-Based on PNN Transformer Fault Diagnosis)

摘　要:提出一种新的基于Pareto多目标进化免疫算法(PMEIA)。算法在每一代进化群体中选取最优非支配抗体保存到记忆细胞文档中 同时引入Parzen窗估计法计算记忆细胞的熵值,根据熵值对记忆细胞文档进行动更新,使算法向着理想Pareto最优边界搜索。此外,算法基于点在目标空间分况进行克隆选择,有利于得到分布较广的Pareto最优边界,且加快了收敛速度。与已有算法相比, PMEIA在收敛性、多样性,以及解的分布性方面都得到很好的提高。(Abstract:This paperproposed a new pareto-based multi-object evolutionary mi mune algorithm(PMEIA). PMEIA selected optmi alnon-dominated antibodieswhichwere then reserved inmemory cellarchive, and introducedParzenwindow to calculate entropy ofmemory cells. Updated thememory cell archive according to entropy ofmemory cells. This guarantees the conver- gence to the true Pareto fron.t Moreover, the performance of clone selection was dependent on distribution in the objective space, whichwas favorable forgetting awidely spreadPareto frontand mi proving convergence speed. Comparedwith the exis- ted algorithms, the obtained solutions ofPMEIA havemuch betterperformance in the convergence, diversity and distribution.)

BP neural network trained by PSO algorithm