Model Predictive Control Algorithm For Double-hormone Artificial Pancreas Based On Three-way Decision Theory

At present,double-hormone artificial pancreas is a research hot topic in the treatment of diabetes,but there are few studies on the switching mechanism of the subsystems for double-hormone artificial pancreas.To solve this problem,a model predictive control algorithm for double-hormone artificial pancreas based on three-way decision switching rules is proposed according to the three-way decision theory and model predictive control algorithm.In order to reduce the dependence on the prior knowledge of the application domain and further improve the stability of the control system,a model predictive control algorithm for double-hormone artificial pancreas based on optimal three-way decision switching rules is proposed.And in the proposed algorithm,the mixed fruit fly algorithm is used to optimize the switching thresholds.The specific research work is as follows:First,the controllers of the insulin infusion subsystem,the zero infusion subsystem and the glucagon infusion subsystem are designed by using the model predictive control algorithm.The prediction models of the three subsystem controllers all adopt the mechanism model,the corresponding objective function is designed and solved with the particle swarm optimization algorithm,so as to achieve the control of the human blood glucose level.In order to further improve the effect of blood glucose control,the three subsystems adopt a combination of interval control and set value control.The blood glucose is first controlled within the normal range and then gradually approaches the set value.Secondly,in order to solve the problem that there are few studies on the switching mechanism of the double-hormone artificial pancreas subsystem,a three-way decision switching rule is proposed.The model prediction controllers of three subsystems for the double-hormone artificial pancreas correspond to the positive,boundary and negative domains of the three-way decision theory,the cost loss function is designed and quantified by the decision rough set theory,and the three-way decision switching rule is determined according to the bayesian criterion.The simulation of 33 virtual patients is carried out with the UVa platform,and the experimental results show that the model predictive control algorithm for double-hormone artificial pancreas based on three-way decision switching rule takes into account the control effect and economic cost,and significantly improved the group control quality of multiple patients.Finally,aiming at the problem that the design of subsystem switching rules for the double-hormone artificial pancreas relies on the prior knowledge of the application field,a optimal three-way decision switching rule is proposed.The total decision risk loss function is designed by using the three-way decision theory,the optimal switching threshold of the three-way decision is optimized by the mixed fruit fly algorithm,and the optimal switching rule of the three-way decision is determined by the bayesian criterion.Simulation experiments are carried out on 33 virtual patients using UVa platform,and the experimental results show that the model predictive control algorithm for double-hormone artificial pancreas based on optimal three-way decision switching rule avoids the dependence on prior knowledge,and significantly improved the group control quality and system stability of multiple patients.