Research On Energy Management Strategy Of Rechargeable Sensor Networks In Power Internet Of Things

Wireless sensor networks(WSN)are the cornerstone to realize the large-scale environmental sensing of Power Internet of Things(PIoT),which has been widely used in many application scenarios such as the source,network,load and storage of Smart Grid.However,the lifespan of sensors is usually limited by the volume.Therefore,energy management combined with wireless energy transfer has always been a key research topic in rechargeable wireless sensor networks(RWSN).For its energy efficiency optimization involving wireless communication,wireless energy resource allocation and the communication-recharging synergy between these two processes,the main contributions of this thesis are as follows:(1)Aiming at the energy efficiency optimization of small-scale RWSN in the PIoT,it is of great significance to allocate energy reasonably by the single mobile wireless charger(MWC).In addition,due to the wireless charging power attenuation,the selection of high-efficiency recharging locations for multiple nodes is crucial to improve the wireless charging utility.Hence,a payoff-maximization-based adaptive hierarchical wireless charging algorithm is proposed.Firstly,the wireless energy allocation process is modeled as a charging payoff maximization problem.Then,using the hierarchical decomposition method,the problem is decomposed into the optimal energy allocation layer based on convex optimality theory,anchor point optimization layer based on the greedy strategy and time slices allocation layer based on the gain recall mechanism.Finally,the energy can be allocated adaptively through the linkage solution among these three layers.The simulation results show that the proposed algorithm can adaptively weigh the effective amount and equilibrium of energy supplement,so as to realize energy efficiency optimization of RWSN based on the single MWC charging.(2)Aiming at the energy efficiency optimization of large-scale RWSN in PIoT,it is of great significance to work together reasonably by multiple MWCs.Therefore,a deep-reinforcement-learning-based multi-device efficient cooperative wireless charging algorithm is proposed.Firstly,the charging payoff function and corresponding optimization problem,which is decomposed into two layers by the hierarchical decompose method to be solved efficiently,are proposed.Then,in the bottom layer,the optimal energy allocation of the single MWC within a given charging sector is realized by convex optimality theory.Finally,based on the feedback from bottom layer,dynamic charging sector allocation is realized by the deep reinforcement learning based on twin delayed deep deterministic policy gradient,where multiple devices are collaborated according to the optimization objective.The simulation results show that the proposed algorithm can ensure the optimal energy allocation of the single MWC.On this basis,charging sectors can be allocated reasonably for multiple devices,so the energy efficiency optimization of RWSN based on multiple MWCs charging can be realized.(3)In the RWSN where the fixed base station can’t cover each node,some of the nodes can’t directly communicate with the base station due to the limit of communication distance,and can only upload data through multi-hop relay.How to significantly optimize the communication-recharging synergy energy efficiency while ensuring the quality of service(QoS)of the network is crucial.Therefore,an energy-efficient multi-hop routing algorithm based on lifespan expectancy balance is proposed.Firstly,a multi-hop routing considering delay is designed.Then,joint recharging scheme,the energy state of nodes can be accurately quantificated by the expected life quantification method considering relevant influencing factors.Finally,a global relay range optimization method based on the chain gain contraction mechanism,aiming at maximizing the minimum lifespan expectancy,is designed to improve the network energy efficiency.The simulation results show that the chain gain contraction mechanism can increase the minimum lifespan expectancy combined with charging scheme on the premise of ensuring the QoS of the network,so as to realize the communication-recharging synergy energy efficiency optimization of RWSN based on partial-coverage fixed base stations.(4)In the RWSN where the fixed base station can cover all nodes,any node can directly communicate with the base station besides multi-hop relay.In order to optimize the communication-recharging synergy energy efficiency through reasonable switching routing mode of each node,a deep-reinforcement-learning-based adaptive dual-mode energy-efficient routing algorithm is proposed.Firstly,using the lifespan expectancy balance theory,the energy-efficient multi-hop routing is designed based on the forward transmission principle.Then combined with the direct transmission route,the adaptive dual-mode energy-efficient routing is designed according to the relationship between the single node’s own lifespan expectancy and the average lifespan expectancy of the whole network.Finally,the deep reinforcement learning method is used to enable nodes to get the above lifespan expectancy relationship without gathering large-scale network state information.Simulation results show that the proposed algorithm can dynamically switch the routing mode of each node according to the energy state.Using deep reinforcement learning,corresponding nodes can achieve up to a high correct rate of routing mode selection according to the incomplete network state information,so as to realize the communication-recharging synergy energy efficiency optimization of RWSN based on the full-coverage fixed base stations while ensuring the applicability.(5)For the multifunctional integrated mobile platform with MWC and base station,how to solve the energy efficiency coupling problem between wireless charging and data gathering is particularly critical.Therefore,a synergy-payoff-maximization-based rechargeable mobile data gathering algorithm is proposed.Firstly,an independent-lifespan-expectancy-balance-based energy-efficient mobile data gathering strategy is proposed.Then,considering energy supplement and consumption,the synergy payoff function and corresponding optimization problem for maximizing the synergy payoff is established.Finally,the problem is decomposed into two layers according to the hierarchical decomposition method.The optimal recharging time allocation is realized based on convex optimality theory in the bottom layer.In the top layer,the optimal working scheme with the highest synergy payoff can be established according to the feedback from bottom layer.Simulation results show that the proposed algorithm can reasonably collaborate wireless charging and data gathering,so as to realize the communication-recharging synergy energy efficiency optimization of RWSN based on mobile base station.