Research On Autonomous Device Cloud Service-Oriented Large-scale Simulation Modeling Of Hydropower Plants

With a large number of large-scale power plants such as the Three Gorges have been built and put into operation and the intelligent transformation of traditional power plant,as well as the implementation of the national "The Belt and Road" hydropower equipment "go out" strategy implementation,the simulation system,as an ideal platform for hydropower operation training and analysis,has extensive and urgent demands.Mass customization and on-demand access in the Internet environment to the hydropower simulation system is a key problem to be solved urgently in the field of the hydropower production in china.The hydropower operating environment is a typical complex large-scale system involved in hydraulic,mechanical,electric,computer control and other various fields,with a large scale,a large number of equipment,and strong coupling between dynamic physical dynamic process and discrete information process.The interaction between "man-machine" and the equipment is intelligent and autonomous,and the working condition is difficult to be arranged in advance.The traditional method of modeling hydropower large system logic decomposition and simultaneous equations takes little account of the new features and new demands of the large-scale simulation and intelligent operation and thus is difficult to solve the new problems of model decoupling and integration of cloud,distributed interaction and coordination,high concurrent expansion and model fault-tolerance in the large-scale customization development of hydropower simulation system under the new cloud simulation application environment.This paper,from a unique perspective of intelligent equipment operation of hydropower station,uses modeling the intelligent and autonomous characteristics of hydropower devices as the starting point,with the combination of cloud computing technology,to search for the modeling theories and methods that match large-scale simulations of hydropower stations and service models of "Internet +hydropower simulation".By given intelligent perception and active service ability,the models are the same as the real intelligent hydropower devices in organization and interaction,so as to make the system decoupling and integration in the way of conforming to the operation mechanism of the real hydropower plant.Additionally,on this basis,the paper seeks for solutions of distributed model,high concurrency,expandable and fault tolerant which match the cloud computing environment.By setting up a hydropower station cloud simulation platform and operating structure that accommodates with the era of"Internet +",this paper provides a new modeling theory and an implementation plan for practical problems of major large-scale hydropower station simulations projects in the field of hydropower production which need to be solved urgently.The main research contents and innovative points are as follows:1)Focus on the new intelligent features and new large-scale simulation demands of hydropower plants,a new modeling theory and computer model known as autonomous device-oriented cloud service adapting to the large-scale simulation and autonomous service synergy are proposed.This method deems the hydropower device as the organization unit of decoupling and simulation service.By modeling the intelligent perception and active service of hydropower device,and implanting the service combination and interaction ideas of SOA and EDA into the intelligent hydropower device models based on the cloud message bus,we can construct the model architecture of loose coupling organization,distributed cooperative and asynchronous non-blocking parallel interaction with the same structure and mechanism of real hydropower station,so as to develop a new "ADaaS"(autonomous device model as a service)cloud service model for large-scale simulation of hydropower stations on the basis of the general SaaS cloud service model.2)It puts forward an autonomous device-oriented cloud service modeling method and framework.It defines and implements core components of the frame,including active perceivable device models,perceived messages,virtual bus,and model containers,etc.In terms of design,it makes a break-through of the currently popular stateless service pattern,i.e.the separation of service from state data in service architecture;realizes the innovation of "In-Memory" mode that holds state data in the device model service.This can greatly improve model’s processing speed and ensure the real-time simulation calculation.3)By the inversion of decision-making of model distribution,innovatively transfer the message routing of the distributed messages publish-subscribe into a classic problem of model coordination and distribution in the cloud computing resource note.By this way,a lot of classic intelligent optimization algorithm can be used for modeling and solving the problem,and a improved PSO algorithm and the optimal solution are given based on catfish effect.At the same time,the distribution of autonomous device cloud service model can be determined by message routing,which enables the service model be equipped with horizontal spreading and fault-tolerant migration ability under the cloud platform.Thus,it provides "cheap" solutions for the high-concurrency,high-availability characteristic of large-scale simulation from a software model level,and can exploit the advantage of cloud computing.4)In accordance with strong coupling features of main electrical system for hydropower production,module alternating segmentation algorithm is put forward,providing algorithm basis for the application of loose coupling autonomous device model in the strong coupling hydraulic-mechanical-electric dynamic process.The graphical power network topology search algorithm is proposed and has been applied to power flow calculation and fault calculation of power system.Based on the thought of alternating segmentation algorithm,units and electric network simulation model for each subsystem and equipment and its dynamic process are built.5)In accordance with numerous complex problems of auxiliary system devices,the idea of hierarchical modular modeling is put forward.The idea is to divide the system into four levels:interactive level,process level,component level,and system level.By virtual device abstraction after the system is divided,it could effectively reduce the number of autonomous device models.By extending the topology search algorithm,which is based on the fusion of connecting lines,to flow net system,the modeling method of general flow net dynamic equation is given.In the end,the modeling method and the framework of autonomous device cloud service is applied to the hydropower simulation cloud computing platform of a certain large power group hydropower training center,which has proved the effectiveness of the method and the framework.