Algorithms For Charging Trajectory Optimization In Wireless Rechargeable Sensor Network

In recent years,wireless sensor network has been widely used in many fields such as national defense and military,intelligent transportation,environmental monitoring.Due to the maturity of wireless charging technology,the wireless rechargeable sensor network(WRSN)has gradually replaced the original network structure,by introducing charging equipment and designing charging strategies to extend the life of the network.Therefore,designing reasonable charging strategies for the network and effectively using the energy of charging equipment have become current research hotspots.Most of the existing works use mobile charger(MC)devices to charge the wireless sensors in the network,but the proposed charging strategies often does not consider the real-time energy consumption of WRSN caused by energy heterogeneity,so that it cannot meet the charging requests of some sensors,this will seriously affect the lifespan of WRSN and the energy utilization rate of MC.To solve the above problems,this thesis further considers multiple factors that affect charging efficiency on the basis of the traditional charging planning problem,including the requirement of multi-cycle charging,the limitation of real-time sensor energy consumption,and three-dimensional space path planning.Specifically,this thesis introduces a charging unmanned aerial vehicle(CUAV)to replace the MC traveling on land,the NP-hardness of the new problem is proved,and three novel charging path optimization algorithms are devised:For the multi-period charging problem,this thesis proposes energy consumption filtering mechanism(ECFM)and priority sorting mechanism(PSM),and design an NSGA-II-based multi-objective charging planning algorithm(MOCP).The ECFM can find effective nodes in each charging cycle and coded them for subsequent sorting.The PSM takes the maximization of charging efficiency and the minimization of the death rate of sensor nodes as two optimization goals,and calculates the priority of nodes and reorganizes individuals to provide additional solution angles.Thanks to the above two mechanisms,the algorithm can avoid falling into local optimal solution as much as possible when providing the charging path in a single round cycle,and select the optimal solution from non-inferior solution set as the final charging path in this round cycle.Experimental results show that the proposed algorithm can effectively reduce the death rate of sensor nodes,and thereby prolong the working life of WRSN.In order to further improve the computational efficiency,this thesis proposes an online algorithm for charging path planning named attention-based mobile charging optimization(AMCO).This algorithm can incorporate energy consumption and time constraints into the multi-objective optimization algorithm,thus it can dynamically adjust the weight coefficients related to the charging state according to the changes in the network environment,and filter out the effective sensor set in each charging period.This algorithm can also dynamically adjust charging frequency of CUAV to wireless sensors according to the size of the remaining power.Experimental results show that the AMCO algorithm can significantly reduce the death rate of sensor nodes and ensure the charging efficiency of MC under different network scales.Different from the previous two fixed-height charging path optimization algorithms,this thesis further analyzes the impact of three-dimensional space path planning on charging efficiency,establishes a “single source and multiple receiving units” wireless energy transmission model,and propose a dynamic stereo spatial route planning algorithm(DSSRP).On the basis of the AMCO algorithm,the algorithm can continuously optimize the hovering position of CUAV in the three-dimensional space using the minimum node set coverage and the minimum circle coverage mechanism,and select the optimal subset solution which satisfies the constraints of charging energy consumption to minimize the energy loss of charging sensor nodes under a reasonable charging height.Comparative experiments prove that the proposed algorithm can further improve the robustness of WRSN and the average remaining power of the nodes.