Research On Demand-side Situational Awareness And Orientation Under Energy Internet

Under the goal of carbon neutrality and new power system construction,China’s energy and power industry has entered a period of transformation and upgrading with the core of cleanliness.The large-scale integration of intermittent renewable energy,represented by wind and solar power,poses significant challenges to the safe and stable operation of power systems.Fully exploiting the potential of demand-side resources in promoting the safe and efficient consumption of renewable energy has become one of the important ways to ensure the high-quality development of energy and power industry.At the same time,with the vigorous development of Energy Internet,and the popularization and application of digital information technology,the integration of digital revolution and energy revolution continues to deepen.The intelligent transformation of energy provides favorable conditions for the development of on the demand side,includes development of multiple resources,the emergence of new entities and new business forms,and technological innovation and transformation and etc.In this context,this paper is mainly guided by the macro strategy of carbon neutrality,new power system construction,and digital transformation,etc.,and focuses on energy production,consumption,operation,and trading within the scope of the demand side,as well as the management of massive energy data that comes with it,this article decomposes the demand side situational awareness and optimization problem into five sub problems:demand side situational awareness functional framework,demand side distributed renewable energy situational awareness,demand side temperature control load resource situational awareness,demand side resource situational orientation,and energy data asset governance,in order to help achieve the goals of the carbon neutrality and the construction of new power systems.The research mainly includes the following aspects.In view of the needs of energy industry transformation and upgrading,on the basis of studying and judging the trend of clean,digital,and quasi centralized development of the Energy Internet,this paper puts forward a technical architecture system of demand-side management in the new era,which aims to crack the energy impossible triangle,takes aggregation operators as the hub,takes information technology as the support,and takes massive data as the production factors.Secondly,on the basis of introducing the concept of situational awareness,a technical hierarchy of demand-side situational awareness under the Energy Internet is proposed,which includes situational perception,comprehension,projection and orientation.Finally,according to the characteristics and demands of demand-side situational awareness.a functional framework of demand-side situational awareness based on cloud-edge collaboration is constructed,and corresponding key supporting technologies are defined,which provides important theoretical support and directional guidance for subsequent research.In response to the issue of demand-side distributed renewable energy situational awareness,a distributed renewable energy situational comprehension and projection model based on two-stage clustering and deep neural network is constructed.Firstly,in view of the influence of temporal and spatial delays on data mining of distributed renewable energy,a feature mining model of equipment operating conditions based on two-stage clustering was proposed on the basis of analyzing the adaptability of dynamic time warping and fuzzy C-means technology,which realized the spatial and temporal reconstruction of historical data and the effective division of operating conditions.Secondly,in response to the modeling and prediction problem of distributed renewable energy equipment under highdimensional information,a situational comprehension and projection model integrating convolutional neural networks and recurrent neural network is constructed to capture and learning of spatial morphology and temporal characteristics of distributed renewable energy output curves.Finally,the case shows that the model constructed in this paper can effectively improve the accuracy of the division of operating conditions and output prediction of-distributed equipment without significantly increasing the calculation cost.To address the issue of demand-side temperature control load resource situational awareness,the baseline power situational comprehension and projection model and evaluation model of temperature control load based on data-mechanism are constructed.Firstly,by introducing Minkowski’s sum theory to build the power model of temperaturecontrol load cluster,the aggregation and selection behaviors of aggregation operators are described,and the defects of traditional sum method are overcome.Secondly,the influence factors and prediction difficulties of baseline power of temperature control load are analyzed,and the comprehension and projection model of temperature control load is constructed by integrating timing neural network and recurrent neural network,in addition.attention mechanism is introduced to improve the learning efficiency of neural network.Thirdly,a performance portrait method of comprehensive regulation of temperature control load was constructed from the perspective of adjustability and reliability.Based on the analysis of the influence of unstable working conditions and other factors on load evaluation,an improved comprehensive evaluation method based on evidence theory is proposed to improve the robustness of evaluation results.Finally,the case shows that,first.the baseline power situational comprehension and projection model constructed in this paper can effectively capture the morphological characteristics and temporal characteristics of baseline power curve in peak-valley period,thus significantly improving the accuracy of prediction results.Second,the evaluation model proposed in this paper can better characterize the comprehensive regulation performance of temperature control load and shield the interference of unstable working conditions.Aiming at the demand-side situational orientation problem under RPS.a demand-side resource situational orientation model driven by reinforcement learning is constructed.First of all,through policy analysis,the logical relationship among RPS,aggregation operators,green certificate trading and other elements is clarified,and the mechanism framework for aggregation operators to carry out demand-side situational orientation under RPS is constructed.Secondly,a demand-side resource situational orientation model and a solution algorithm based on reinforcement learning are proposed,which effectively simulates the continuous decision-making process of aggregation operators in a dispatching cycle.Finally,the example shows that,first,under the RPS,taking the aggregation operator as the hub to carry out demand-side situational orientation can realize the tracking of renewable energy output,so as to increase the income of the aggregation operator while increasing the consumption of renewable energy.Second,the increase of RPS weight and aggregator operators’ preference for long-term benefits can promote the consumption of renewable energy to a certain extent,but this effect is not linearly positive,and further consideration should be given to the influence of factors such as price difference of energy products.Aiming at the problem of energy data governance in Energy Internet,a life-cycle governance model of energy data assets based on data curation theory is constructed,Firstly,this paper analyzes the development of Chinese energy big data from the perspectives of connotation,technology and policies,and points out that the energy data ecosystem is one of the important development trends of energy big data.Secondly,in view of the key challenges facing the development of energy data ecosystem,the concept of data curation is introduced,and a whole-life cycle governance model of energy data assets based on data curation theory is constructed from the perspectives of overall framework,strategic planning,integration,integration and etc.Finally,from the dimensions of laws and regulations,technology application and system mechanism,the paper explores the solutions to key issues in energy data asset governance,such as data rights and data fusion,in order to provide corresponding support for the continuous deepening of the integration of digital revolution and energy revolution.