基于CPSO滚动优化的堆芯功率自适应预测控制

Adaptive Predictive Control for Core Power Based on CPSO Rolling Optimization

针对核反应堆变功率过程中系统的非线性和反应性增量约束问题,本文将具备模型参数自适应辨识能力的改进型广义预测控制器(JGPC)应用于堆芯功率控制,该控制器通过预测模型参数和递推关系计算未来时刻的预测输出值,同时采用经正弦混沌策略和非线性惯性权重改进的混沌粒子群算法(CPSO)进行滚动优化,在优化过程中通过设定优化边界和混沌策略来处理反应性约束;以堆芯功率的受控自回归积分滑动平均模型(CARIMA)作为预测模型,并采用遗忘因子递推最小二乘法(FFRLS)自适应辨识模型参数,以克服堆芯功率模型非线性。基于MATLAB平台对本文控制器进行仿真验证,结果表明,该控制器在满足约束条件的情况下,能使堆芯功率快速、稳定地跟随设定值,且具备一定的抗干扰能力。

In view of the nonlinear and reactive constraint of nuclear reactors in the process of variable power,this paper proposes an improved generalized predictive control(JGPC)for core power control.The JGPC calculates the predicted output value by predicting the model parameters and recursive relationships.At the same time,chaos particle swarm optimization(CPSO),which is improved by the sinusoidal chaos strategy and nonlinear inertia weight,is applied to the rolling optimization of JGPC.In the process of optimization,the reactive constraint are dealt with by setting optimization boundary and chaos strategy.The controlled auto-regressive integral moving average(CARIMA)model of core power is established as the JGPC prediction model,and the forgetting factor recursive least squares(FFRLS)method is used to identify the model parameters online.The JGPC controller is simulated and validated based on MATLAB platform.The results show that the controller can make the core power follow the set value quickly and steadily under the condition of satisfying the constraint,and has a certain anti-interference ability.