For the purpose of addressing the difficulty of improving load forecasting accuracy brought by enormous input data features, a method based on hybrid neural network using parallel multi-model combination is proposed. In order to respectively extract local features and time-series features, this paper places the convolutional neural network (CNN) in parallel with the gated recurrent unit (GRU) structure, then concatenates the output of two network structures and inputs to a deep neural network, uses deep neural network to perform load forecasting. Through a prediction experiment of load and temperature data by using the proposed method, the experiment results show that, compared with GRU-NN model, long short term memory (LSTM) model, serial CNN-LSTM network model and serial CNN-GRU network model, the proposed method shows better prediction performance.
In addition to solving the problems of on-site photovoltaic and wind power and its stable operation, the optimization scheduling strategy for microgrid should also have flexible resources such as distributed power and loads with demand-response capabilities to provide auxiliary services to the grid and capabilities to participate in real-time scheduling of the upper-level grid. This paper proposes a microgrid resource optimization scheduling strategy based on BP neural network. This paper combines the microgrid operating costs and demand-response capacity gains to establish an economic optimal scheduling strategy for the day-to-day stage. The intra-day simulation stage simulates the forecast of power fluctuations and the real-time demand of the upper-level grid, and learns through the neural network to obtain the intra-day scheduling model to prepare for intra-day scheduling. In the daytime phase, through the demand-response signal of the upper-layer power grid, the power of the contact line is input into the training model of neural network, and the real-time power of each distributed power source in the daytime phase is obtained. The strategy proposed in this paper can not only guarantee the economic operation of the microgrid, but also meet the real-time dispatching requirements of the upper-level grid. Finally, the economics and effectiveness of the strategy are verified by the results of day-after optimal scheduling examples.
In order to ensure the proportion of renewable energy in power consumption in provincial regions, the government has introduced an assessment system for the responsibility weight of renewable energy consumption, which transfers the assessed party from the generation side to the user side. The change of assessment method will greatly affect the interests of thermal power producers and their enthusiasm to participate in tradeable green certificate (TGC). Thermal power accounts for a relatively high proportion of electricity generation in China. Whether thermal power producers respond positively to the green power market will greatly affect the implementation effect of the new policy. The government needs to set reasonable policy parameters to guide the trading strategies of thermal power producers so as to achieve the ideal equilibrium result of the game. According to the new changes, this paper constructs an evolutionary game theory (EGT) model that combines green power producers, thermal power producers and assessed users under bounded rationality, and simulates the evolution of thermal power producers’ strategies and the changing trend of TGC market under different parameters. The results show that, under the existing conditions, all thermal power producers will reach the ideal evolutionary stable strategies (ESS) point to participate in TGC trading within the policy parameter fluctuation of ±10%. However, when the policy burden increases, the rate of thermal power producers to reach ESS point will slow down significantly. The government should maintain or appropriately reduce the current policy burden level, and encourage thermal power producers to actively respond to TGC transactions, so as to ensure the efficient and stable implementation of the new policy.
To promote the accommodation of renewable energy, when renewable energy power generation enterprises participate in the direct purchase of Shanghai power market, all their power needs to be purchased. To satisfy this policy, some provinces use a trading mechanism called secondary clearing, which lacks of relevant research. Therefore, it is necessary to study the transaction process and pricing method of this mechanism. According to the power market model of bilateral bidding and marginal clearing price, this paper establishes a mathematical model of power market which combines matching clearing and secondary clearing, puts forward three principles that pricing of secondary clearing should conform to, and analyses the problems existing in marginal price. A new pricing method is put forward by illustrating with an example.
With the development of distributed generation and DC loads, a safe, reliable, and efficient AC-DC hybrid distribution network will become an important form for future development of distribution network. The DC distribution network based on the voltage source converter has the advantages of strong controllability, few power quality problems, and good economy. At the same time, it also introduces a problem of power flow calculation that is different from the AC distribution network. This paper firstly establishes a model for power flow calculation that takes into account inverter losses and DC converter losses, and gives calculation method for loss parameters. Secondly, the method for solving the DC power flow model is introduced. When analyzing the power flow of the DC system, the influence of control method of the voltage source converter and the DC / DC converter on the power flow calculation is mainly considered. Different power flow solving methods are given for different control strategies. Several effective solutions are given when the flow exceeds the limit. Finally, an IEEE 33-node DC distribution network example is used to verify the correctness of the loss model proposed in this paper and the effectiveness of the power flow solution method.
At present, there is a lack of unified and standardized typical examples for the scenario that distributed generation is widely connected to the distribution network in the future. Therefore, combined with the actual situation of China’s distribution network, three typical scenarios of rural, urban areas and industrial park are constructed. On this basis, the design principles and methods for typical scenarios are proposed. According to the load type, typical characteristics of distribution network, and the penetration rate of distributed generation, the distribution network examples in typical scenarios are designed, respectively, and the detailed parameters and typical daily data are given. At the same time, the analysis index and method for the example are given, and the detailed analysis and test are carried out from the perspective of over / low voltage test and distributed generation accommodation calculation. The results show the rationality and feasibility of the design example, which can provide a good case basis for the related research of distributed generation access.
With the increasing penetration of renewable distributed generators (DGs), combined with the power regulation capabilities of soft open point (SOP) and the characteristics of energy storage system (ESS), SOP integrated with ESS (ESOP) can adjust the power flow and enhance the economy of active distribution networks (ADNs) in spatial and temporal aspects. Considering the complex physical structure and high investment cost of ESOP, reasonable capacity configuration of inverter and energy storage battery should be selected in the planning stage to enhance the comprehensive benefits of the ADNs. Firstly, the structure and mathematical model of ESOP are elaborated in this paper. Secondly, to establish the ESOP planning model of ADNs, the minimum comprehensive cost of the network is taken as the object and the operation constraints of ESOP and ADNs are considered. Finally, the second-order cone-programming algorithm is used to solve the proposed model, and the effectiveness of the proposed planning model is verified by a modified IEEE 33-node test system. The results show that the proposed ESOP planning model can effectively reduce the comprehensive cost and improve the economy of ADNs.
The large-scale development of wind power resources and the increasing proportion of natural gas consumption in primary energy have brought new challenges to the planning and operation of power systems. In order to transmit large-capacity wind power to load centers thousands of kilometers away, UHV DC transmission systems with long-distance, large-capacity, and low-loss power transmission capabilities are adopted. This paper proposes a statistical analysis method of wind-gas transmission capacity based on the confidence level for long-distance transmission of energy bases with abundant wind power and natural gas resources. Considering comprehensive various factors such as randomness of energy output, economic, technical and environmental nature of transmission scheme, and endurance capacity of receiving end system, a long-distance DC transmission evaluation index system is established for the wind-gas energy base. According to improved prospect theory, candidate schemes for HVDC transmission systems are comprehensively optimally selected. In the example analysis, the application scope of different transmission schemes is analyzed by the sensitivity method, and the superiority of the proposed planning scheme is verified, which provides a reference for scientific decision-making of large-capacity long-distance DC transmission planning.
The development of metrology communication technology enables the user’s load information to be accurately collected, and the clustering analysis of the power consumption characteristics of the load can be performed. In order to solve the problem that load clustering application scenarios need clustering results as similar as possible to initial cluster center, two improved ant colony clustering algorithms are designed on the basis of the ant colony clustering algorithm. The two factors that determine the clustering effect and the distance between the cluster center and the initial cluster center constitute the fitness index instead of the traditional mean square error for updating pheromone matrix. The example analysis shows that the algorithm can solve this kind of application scenario well and has good clustering effect.
In order to study the temperature rise characteristic of the saturable reactor for UHVDC converter valve and obtain the temperature distribution laws of the cores inside, this paper analyzes the calculation models of saturable reactor loss and core temperature rise on the basis of existing researches. A temperature rise test mothed for saturable reactors is proposed on the basis of embedded optical fibre temperature sensors, by which the temperature of all the cores inside could be measured directly. Two test conditions are designed considering the new generation of ±800 kV/8 GW UHVDC transmission project, including the equivalent core loss loading method by high frequency power source and valve section operating test loading method by synthetic test circuit. The temperature of the key points of four kinds of UHVDC valve reactors are measured and the temperature distribution characteristics of cores inside are analyzed comparatively. The results show that, under two test conditions, the temperature of the cores close to cooling water inlet and outlet are lower than others in one valve saturable reactor, and the temperature rise characteristics of each type of saturable reactor are basically consistent. Under operating condition, the core temperature of each type of valve reactor is lower than 90℃ and the case temperature is lower than 75℃, which can meet the requirements of UHVDC project.
The power retailers face the risk of electricity deviation penalty in the transitional period before the formation of the spot market of electric power in China. In this paper, the CCHP system is provided as an effective means to reduce the deviation power because of its controllable output. At the same time, the rolling optimal model of the CCHP considering the cost of deviation power penalty is proposed under the daily deviation penalty mechanism, which solves the operability problem of reducing the deviation power. The simulation is performed on the basis of the load of a power retailer. The simulation results show that, no matter in the case of positive deviation or negative deviation, by rolling optimization of the operation mode of CCHP, the cost of deviation penalty can be reduced and the overall revenue of the power retailer can be improved.
Clean energy power accommodation is a key goal that China has been chasing currently since the power market reform was carried out. Under this background, this paper constructs a peak-regulation optimization model for gas-fired generators in parks with power-to-gas units employed under the mixed market environment. Firstly, the structure of a park power system with photovoltaic and wind power generators, gas-fired generators, a power-to-gas converter, and a gas storage tank is designed. Secondly, with consideration of the green certificate market and peak-regulation compensation mechanism, green certificate trading models for power units in the park and a peak-regulation compensation model are both built. Thirdly, a mathematical model of economic dispatch for park power supply systems with power-to-gas units and clean energy power generation is built, in the pursuit of the minimum CO2 emission and the maximum clean energy power accommodation, peak-regulation willingness of gas-fired generators and system’s revenue. The model is solved by using the chaos particle swarm optimization algorithm. The following results are obtained: 1) the consumption rate of clean energy power has increased 6.69% and the CO2 emission has reduced 14.66 tons, in the scenario with both power-to-gas and peak-regulation compensation mechanism involved; 2) the revenue of the system has increased up to 187,060 Yuan, after fully considering gas-fired generators’ participation in peak regulation, thus verifying the effectiveness of the proposed model and showing a better convergence effect of the algorithm.
When renewable energy such as photovoltaic and wind power replace synchronous generators in the grid, the inertia of the grid and the primary frequency-regulation reserve capacity will be reduced, and the transient stability of the grid frequency will also deteriorate. According to the analysis of the relationship between the coefficient of energy storage controller and the frequency dynamics of power grid, a capacity allocation method for energy storage is proposed. The method can maintain the system inertia and primary frequency-regulation capacity constant before and after renewable energy replace the traditional units. Meanwhile, considering the fluctuation of the output power of the renewable energy, the load supplied by renewable energy is lower than its installed capacity. To improve the economy of energy storage allocation, this paper further proposes a method to determine the energy storage capacity by setting the confidence level according to the historical output power data of renewable power stations. Simulation results show that the proposed allocation method for energy storage capacity can quantitatively compensate the inertia and reserve capacity of the power grid, thereby effectively improving the transient stability of the power grid frequency.
Due to its variable-speed constant-frequency operation characteristics, DFIG has become the current mainstream wind power generators. DFIG can generate and absorb reactive power, and support the voltage of point of common coupling (PCC). On the basis of the analysis of the equivalent circuit of DFIG, according to the relationship between the active and reactive power output from the stator side of DFIG, a coordinated voltage control strategy of DFIG under power-limited operation and SVG is proposed. Under normal condition, DFIG operates in Maximum Power Point Tracking (MPPT) mode, and SVG controls the voltage of PCC within a reasonable range. When the voltage of PCC exceeds the limit, DFIG enters the power-limited operation mode. Then DFIG and SVG coordinately control the voltage of PCC and the reactive power regulation capability of DFIG is in priority use. Finally, the effectiveness of the proposed voltage coordinated control strategy is verified by simulation.
With the continuous growth of the installed capacity of wind power, the wind power accommodation becomes a key factor in power gird operation. According to the concept of the transactive energy systems which has been widely investigated in academic area, this paper proposes a framework of wind power utilization by using a series of trade contracts based on transactive energy control (TEC) method. Furthermore, the smart contract technology with TEC is jointly considered in this paper. The trade logics proposed by participants are formulated to a series of smart contracts through the assisted smart-trade decision-making method, and finally the control objective of wind power utilization is achieved. The IEEE 30-node system is used to verify the effectiveness of the proposed control method.
