Search

This paper is intended as a companion to [12] and focuses on the numerical implementations of the method in the context of insurance problems with mixed controls: reinsurance acting as continuous control, and capital injection and dividend payment acting as singular ones. The aim is twofold. On the one hand, we provide a comparison of the output of our method in terms of the optimal value against two benchmarks. The first benchmark is provided by [4] in the no-reinsurance setting and when the claims are exponentially distributed. The second benchmark refers to [8] where capital injections are not permitted, but different claim distributions can be employed. On the other hand, we illustrate the improvement of the models in [8] when capital injection is allowed. We also look into the effect of reinsurance in the model [4]. Several cases for the equity cost parameter are discussed as are the boundary (or limit) situations. • We provide a validation of the Dual Linear Programming algorithm by showcasing its performance on two benchmarks. • With a relatively modest order of moments (4), the relative errors are of at most 3%. • Leveraging this algorithm reveals significant enhancements in the expected dividends, advocating for capital injections. [ABSTRACT FROM AUTHOR]

This paper focuses on linearisation techniques for a class of mixed singular/continuous control problems and ensuing algorithms. The motivation comes from (re)insurance problems with reserve-dependent premiums with Cramér-Lundberg claims, by allowing singular dividend payments and capital injections. Using variational techniques and embedding the trajectories in an appropriate family of occupation measures, we provide the linearisation of such problems in which the continuous control is given by reinsurance policies and the singular one by dividends and capital injections. The linearisation translates into a dual dynamic programming (DDP) algorithm. An important part of the paper is dedicated to structural considerations allowing reasonable implementation. We also hint connections to methods relying on moment sum of squares and LMI (linear matrix inequality)-relaxations to approximate the optimal candidates. [ABSTRACT FROM AUTHOR]

• A new kind of data-based fault detection method is proposed, which combines adaptive dynamic programming and multiple-modeling. • A stable working state is established by using ADP and history data. Once the system runs out of the stable state, it's known for sure that fault occurs. • A whole stable working space of the system can be built with the method of multi-modeling and data of all working modes, as long as system states remain in the working space, system is working normally, exclude the disturbance of modes switching, and improve the accuracy of the fault detection algorithm. • The fault can be isolated, if the direction and magnitude of the deviate vector are solved, which can be achieved by approximation and iteration using the method of ADP. This paper presents a adaptive dynamic programming-based fault detection and isolation (FDI) scheme to detect and isolate faults in an aircraft jet engine. To this end, the weights in Actor-Critic neural networks are first tuned to learn the input-output map of the jet engine considering its multiple working modes. The convergences of the trainings in Critic-Actor neural networks are strictly proved without knowing the drift dynamics and the input dynamics in the presence of unknown nonlinearities and approximation errors. Using the residuals that are generated by measuring the difference of each network output and the measured engine output, various criteria are established for accomplishing the fault diagnosis task, that addresses the problem of fault detection and isolation of the system components. A number of simulation studies are carried out for combustion chamber of a single-spool jet engine to demonstrate and illustrate the advantages, capabilities, and performance of our proposed fault diagnosis scheme. [ABSTRACT FROM AUTHOR]

• This article develops an online ADP method for discrete nonlinear systems, it can reduce the computation burden. • Two ESNs are employed to implement the online learning and the output weights are obtained simultaneously. • The hidden layer of ESN is generated randomly rather than designed with much effort, which reduces the application complexity. This paper investigates the online optimal control problem of discrete-time nonlinear systems using echo state network (ESN)-based adaptive dynamic programming (ADP) method. An online iterative learning algorithm is proposed to solve the partial differential Hamilton–Jacobi–Bellman (HJB) equation in real time. A novel neural networks (NN) critic-actor architecture is presented using two ESNs to implement the ADP method. Then, two online learning laws of the output weights are designed for searching the optimal cost function and control policy. The stability of system and output weights is analysed using Lyapunov approach. Three simulations are given to show the feasibility and effectiveness of the designed algorithm. [ABSTRACT FROM AUTHOR]