Coordinated Optimal Control For The Steam/Water Loop In Large Scale Ships' Steam Power Plant

Steam power plant can gain larger amount of power with smaller size and lighter weight,and the vibration in the steam power plant is smaller compared with other power plants.Hence,there are many large scale ships apply steam power plant,including Liaoning aircraft carriers.However,the structure of the steam/water loop in large scale ships is complex,and there is a large number of equipment in the steam/water loop.The sub-loops in steam/water loop are strongly coupling with each other.There are multiple stable operating conditions and dynamic conversion processes during the operation of the steam power plant.Also,the operating conditions change a lot and the disturbance is frequent.In order to ensure stable control of the steam/water loop in large scale ships and improve its control performance this paper conducted and completed the study about coordinated optimal control for the steam/water loop in large scale ships.Firstly,in order to obtain the mathematical model of the steam/water loop in large scale ships,the system structure and operation mechanism of the steam/water loop were analyzed,and the system is divided into three sub-loops,including supercharged boiler,deaerator,condenser and exhaust manifold.With the method combining the system mechanism and experimental data,each model of the sub-loops is obtained.The accuracy of the system model was verified with load increase and decrease experiments.It laid the foundation for the simulation and control system design of the steam/water loop in large scale ships.Secondly,a smooth sliding mode control strategy is proposed for the drum water level control in the supercharged boiler to deal with the problems such as: parameter perturbation,frequent interference,and non-self-balancing capability.By designing a terminal sliding mode surface,and putting the sign function into the second derivative of the control effort,the chattering is solved.The effectiveness of the smooth sliding mode control is verified in the presence of the system model parameter perturbation.The smooth sliding mode control strategy guaranteed the stability of the drum water level with large load changing.Thirdly,the steam/water loop in large scale ships is a complex system with strong interaction and multiple constraints,and a multiple objective optimization predictive control strategy is proposed.A multi-level optimized distributor is designed and,and the control structure is hierarchically optimized.With the proposed method,the safety control is ensured,and the tracking performance and energy saving are both improved.When the deviation of the system states is small,the multi-objective optimization model predictive control strategy converts the constrained optimization problem into an unconstrained optimization problem,which improves the computing speed of the model predictive control algorithm.By simulation,the effectiveness of the multi-objective optimization model predictive control is verified for the steam/water loop in large scale ships.Then,a fractional order model predictive control strategy is proposed to solve the difficulty in configuration of the weighting factors in the traditional model predictive control.By replacing the integer order cost function with a fractional order one,the multi-dimensional weight factors optimization problem is transformed into two-dimensional fractional orders optimization problem,which greatly reduces the optimization dimension.The results from simulation test show that the fractional order model predictive control not only improves the tracking performance of the steam/water loop,but also enhance the anti-interference ability of the system.Finally,a parameter self-tuning control algorithm was proposed to solve the poor control performance of the existing PID control system in steam/water loop in large scale ships.With a single small sine test around the operating point,the frequency response can be obtained.According to the system performance requirements,such as robustness or phase and gain margin,a forbidden region can be obtained on the Nyquist plane.By searching for the tangent point of the system Nyquist curve at the boundary of the forbidden region,the optimal PID controller parameters can be obtained.Simulation experiments are conducted compared with other PID parameter self-tuning methods,and the results show that model-free parameter self-tuning control algorithm improves the control performance of the system while ensuring the system performance requirements of the steam/water loop.