Research On Grid Materials Warehousing Dispatch Optimization Decision-making Model And Decision Support System

With the rapid development of China's economy,the power industry is bound to carry out a large-scale expansion to meet the growing demand for electricity consumption in the entire society.As the core resource for the development of power grid construction,materials have important strategic positions in such areas as power grid infrastructure construction,operation and production,equipment overhaul and emergency dispatch.In order to realize the lean management of the company,the State Grid Corporation of China proposed the goal of intensive management of materials.The material management process has also changed from simple tender and procurement to supply chain management in the entire process of requirements planning,source procurement,warehousing and dispatching.The stable supply of electric power supplies is crucial to the safe operation of the power grid.The materials supply chain involves more business and the environment is relatively complex and changeable.There are many internal and external risk factors.How to effectively identify and control potential risks is a topic that today's grid company leaders need to face.From the perspective of grid supply chain management,this paper divides the material management process into four major segments:demand planning,supplier selection,storage quota optimization,and emergency dispatch optimization.Using the risk identification model,the risk factors under intensification are explored.Build mathematical models based on intelligent algorithms for risk prediction,decision making,optimization,etc.,and complete dynamic risk simulation in material management,so as to leanly manage materials and reduce potential risks.Finally,a risk decision support system for grid supply and dispatch optimization is designed for the proposed model,and the theoretical model is formed into a system.The main research work of the dissertation includes the following aspects:(1)According to the management objective of intensive chemical assets proposed by State Grid Corporation,the material management risk under intensive management is proposed.Analyze the process of the risky agreement inventory procurement and key data nodes to identify the project attributes that have the greatest impact on the intensive risk.On the basis of this,a provincial power company's agreement inventory requirements,procurement and material allocation process data are taken as examples for analysis,to identify the key data nodes and problems in the agreement inventory procurement.Through the association of historical project database and material acquisition data,a model training sample database was established,a BP neural network model of particle swarm optimization was constructed to forecast the protocol inventory requirement,and the efficiency and performance of the algorithm were analyzed.(2)For the issue of supplier selection,first analyze the status quo of the supplier,the problems in supplier selection,supplier classification criteria,and the relationship between the grid and the supplier.Based on the analysis results,risk assessment indicators were designed.Combined with the subjectivity of the evaluation of power grid company leaders,an interval-valued intuitive hesitation fuzzy risk decision model considering credibility is proposed.Finally,through the model analysis results,measures for the supply decision-making risk management were presented.(3)Studying the construction of power supply warehouse from two perspectives:reserve quota calculation and material node location selection.Aiming at the shortcomings of traditional calculation methods and research methods,a BP neural network model based on the optimization of genetic algorithm was proposed to forecast the initial inventory in the beginning of the month.The model also considers risk factors(demand forecasting errors)and network performance(time and economic costs).The prediction accuracy is better than traditional multiple regression methods,and risk factors can also be flexibly adjusted.In terms of material node location,a power supply network optimization model based on robust optimization is proposed.The model is better when faced with uncertain demand.Based on the reserve quota calculation method and material network optimization method,an efficient and reliable power supply network can be constructed to improve the ability of power companies to cope with risk events.(4)Study the power material network to face the sudden disaster risk scheduling problem.First,we study the simpler business scenario,that is,the multi-nodal collaborative optimization problem,and propose a Q-learning multi-node collaborative optimization model for path decision.This model mainly considers the network life cycle and guarantees the service capabilities of the entire network node,thus ensuring the security and reliability of the material network.Then,consider the issue of multi-node disaster scheduling optimization with limited total amount of supplies,which is also the core issue of emergency response for power companies,propose a model of demand priority for multiple disasters based on D-S evidence theory,and gives a clear prioritization of material supply for multi-nodes at the time of disaster.Next based on the collaborative optimization model and priority ranking model,an emergency material dispatch planning model is proposed.Finally,the decision support system is designed according to the above model of grid supply and storage scheduling.Firstly,the system analysis is proposed,and then the content of system design is put forward,including frame design,function design,module design,human-computer interaction design and database design.On this basis,the key technologies needed for system development were discussed.