| Rotorcraft UAV has been widely used in more and more fields due to its vertical takeoff and landing,simple control and easy hover.However,due to its weak computing power,small capacity,limited power consumption and other problems,the accuracy of battery related parameter estimation of electric rotorcraft UAV is insufficient.Therefore,enhancing its battery management ability through cloud has become the only way for electric UAV(cluster)battery management system.Due to working status,task requirements or external interference,UAV often stops updating data with the cloud,resulting in enhanced and ineffective battery management capabilities.Therefore,this paper takes advantage of the cross-platform architecture of Endcloud to optimize and estimate the battery parameters of the UAV battery management system when the data link is interrupted.At the same time,rich resources in the cloud are utilized to design strategies to realize online monitoring of battery parameters in the whole life cycle of electric rotor UAV(cluster),and the monitoring data is used as the decision-making basis for stakeholders.The specific research content of this paper is as follows:First,the requirements of various stakeholders in typical application scenarios were analyzed,and functional design and index evaluation were carried out to form the overall scheme of the cross-platform battery management system of Endcloud.Taking plant protection and search and rescue as typical scenarios,the requirements of UAV operators and equipment maintenance personnel and other stakeholders were analyzed,and eight core requirements including cluster energy analysis,fault location and prediction were sorted out.Design battery parameters based on requirements,such as cloud full-time online,cluster battery life cycle management and other functions and related indicators;An end-cloud collaborative architecture based on reinforcement learning and extended Kalman filter(EKF)is designed based on the hardware base of end-cloud combination,which is used as the overall design scheme of crossplatform battery management system.Secondly,it discusses the overall strategy of cross-platform collaborative battery parameter estimation,designs terminal algorithm and cloud algorithm,and then carries out simulation verification.The basic process of cross-platform collaborative cell parameter estimation algorithm is introduced and a four-layer BMS end-cloud collaborative architecture is proposed.Based on the model-driven approach,the terminal algorithm conducts modeling and HPPC operating condition testing.Combined with the cloud strategy,the EKF algorithm is used to estimate the battery parameters(SOC),and the process of the cloud-based adaptive EKF highprecision battery parameter estimation method is analyzed.By combining reinforcement learning and EKF method,the cloud optimizes the EKF parameter(measurement error covariance R)through reinforcement learning(DQN)to achieve accurate full-time estimation of SOC.Finally,a simulation platform is built and the accuracy of the algorithm is verified by using open source data set.Finally,according to the requirements and overall scheme of the cross-platform battery management system,the cross-platform battery management system was designed and built,and the system platform verification was completed.To its hardware circuit and driver related design.Based on Huawei Cloud platform,terminal device access module,data storage module,cloud battery management module and cloud battery data visualization module are designed according to requirements.Finally,the cross-platform battery management system platform verification is completed through debugging.The terminal battery management system takes advantage of the dual platforms to continuously improve the terminal computing power and improve the accuracy of the battery status estimation.The open source data set and self-measured battery data were used for simulation experiments to verify the effectiveness of the parameter estimation algorithm of the cross-platform battery management system,which has high application value for the battery management units of electric rotor UAV and other unmanned systems. |
PDF Full Text Request