Research On Internal Model Sliding Mode Control For Marine Boiler Drum Water Level System

Oil fired auxiliary boiler for large oil tanker is one of the important equipment on board,the stable operation of the boiler is the precondition for the safe navigation of the ship.It is of great significance for the safe and economic operation of boiler and ship to ensure the fluctuation of drum water level within the allowable range.Regarding the strongly coupled,non-linear,time-delay system of boiler drum water level,the traditional PID control method has some limitations because of its high dependence on the precise model of the controlled object,and the parameters of controller can not be changed after setting,so it has poor adaptability to the time-delay nonlinear system and can not meet the control requirements well.In this paper,firstly,according to the working principle of oil fired auxiliary boiler for large oil tanker,the mathematical model of drum water level system of the boiler is obtained based on the principle of conservation of mass and the law of conservation of energy through certain simplification and assumption,in addition,the influence of feed water flow and steam load on drum water level system is analyzed.For the time delay and non-linear property of the system,an internal model control method for drum water level system is designed based on the analysis of traditional PID control strategy.Secondly,an internal model sliding mode control(IM-SMC)method for drum water level system based on global non-linear integral sliding surface is proposed in view of the shortcomings of internal model control method,this method combines the advantages of internal model control and sliding mode control,which not only compensates the time delay and enhances the robustness of the system,but also improves the transient performance of the system.For non-self-balancing characteristic of drum water level system,a disturbance rejection controller is designed to eliminate the steady-state error of the system after disturbance.Although the sliding mode control law made system have fast response speed and strong robustness in internal model sliding mode control method,it will cause system chattering.Hence,an internal model sliding mode control strategy for boiler drum water level system based on RBF neural network is proposed in view of chattering defect of sliding mode control law.RBF neural network is used to identify uncertainties of system parameters and external disturbances to reduce system chattering.In order to overcome the shortcomings of RBF neural network learning algorithm,this paper adopts hierarchical genetic structure to hybrid code the structure and parameters of the network and proposes a hybrid hierarchical differential evolution algorithm(HHDEA)to optimize the RBF neural network.The simulation results show that the designed control strategy not only reduces the chattering of the system,but also has the advantages of fast response,strong anti-interference ability and robustness after the optimized RBF network is applied to the internal model sliding mode control strategy for drum water level system.