针对锌电解过程各参数之间耦合严重、能耗高、建模困难,研究了锌电解电流效率与各工艺过程参数之间关系的数学模型,提出了一种改进的粒子群优化算法(IPSO)进行模型参数估计,该算法在粒子失活时,对粒子进行变异或扰动操作.重新激活粒子,避免了算法陷于局部最优解,改善了优化算法性能;以锌电解过程实验数据为样本,采用改进的粒子群优化算法对模型进行参数估计和检验,并与基本粒子群算法和BP神经网络模型进行比较,仿真结果证明了模型的有效性。
To overcome the severely coupling of the parameters, the high energy cost and the difficulty of mechanism model establishment of zinc electrowinning process, a mathematic model about the relationship between current efficiency and technical parameters is researched. An improved PSO optimization is proposed to estimate the model parameters. To improve the performance, it uses a special mutation or perturbation to activate particles when particles lose activity. Experiment sample data of zinc electrowinning process is applied to estimate the model parameters. Compared with the base PSO method and BPNET model, the model is validated by the simulation result.