Microgrid optimal scheduling method taking into consideration system operation risk and user satisfaction

Abstract

Disclosed is a microgrid optimal scheduling method taking into consideration a system operation risk and user satisfaction, comprising: generating a power prediction curve of wind power and photovoltaic new energy output in a microgrid, using a Monte Carlo simulation method to obtain a prediction error range to generate a typical scene set, and by means of a data fitting method, acquiring a wind-photovoltaic prediction error probability density distribution function; using value-at-risk and conditional value-at-risk theories to quantify a microgrid operation risk which is increased by slightly high or low output caused by wind power and photovoltaics; obtaining the scheduling cost of an adjustable load in a microgrid system; taking the minimum sum of conditional value-at-risk cost and microgrid system operation cost as an objective function and a user satisfaction level under a demand-side response as a constraint, establishing a microgrid optimal scheduling model taking into consideration the operation risk and the user satisfaction; and using a genetic algorithm to solve the model, and obtaining an optimal result of grid-connected microgrid system optimal scheduling. The safety, economic efficiency, stability and reliability of the microgrid are enhanced.

Description

Microgrid Optimal Scheduling Method Considering System Operation Risk and User Satisfaction Technical Field

[0001] The present invention relates to the technical field of power grid dispatching, and in particular to a microgrid optimization dispatching method that takes system operation risks and user satisfaction into consideration. Background Art

[0002] Microgrids, due to their environmental friendliness, flexibility, high efficiency, and ability to integrate a high proportion of renewable energy, have become a crucial component of my country's intelligent power development. However, the randomness, volatility, and uncertainty of distributed energy sources such as wind and solar power pose significant risks to the safe, stable, economical, and reliable operation of the system. Research is urgently needed to effectively characterize and quantify the operational risks posed by wind and solar power uncertainty to microgrids. Furthermore, the large-scale integration of renewable energy sources has led to difficulties in peak regulation in microgrid systems, and flexible regulation resources on both the power supply and grid sides are relatively scarce. This inevitably poses significant risks to the optimized scheduling of microgrid systems. Fully exploring the flexible regulation capabilities of microgrids on the load side can effectively reduce system operational risks and enhance risk management capabilities.

[0003] Current research on the economic optimization and dispatch of microgrids primarily addresses wind and solar power uncertainty through stochastic optimization and robust optimization. However, these approaches have not been translated into risk costs and incorporated into the economic models of microgrid operation. However, the uncertainty risk of renewable energy output can have a certain impact on microgrid operation, so risk costs should be included in economic models. While microgrid-controlled flexible loads can effectively improve system flexibility, they incur certain dispatching costs and can also alter user electricity consumption habits, impacting their experience. Existing research has not comprehensively considered the impact of demand-side response on the economic efficiency, flexibility, and user satisfaction of microgrid operations.

[0004] Summary of the Invention

[0005] The technical problem to be solved by the present invention is: the present invention provides a microgrid optimization scheduling method that takes into account system operation risks and user satisfaction, so as to solve the problem that although flexible loads accepting microgrid regulation can effectively improve system flexibility, it will bring certain scheduling costs and will also change the user's own electricity consumption behavior habits, affecting the user's electricity consumption experience. Existing research has not comprehensively considered the impact of demand-side response on the economic efficiency, flexibility and user satisfaction of microgrid operation.

[0006] The technical solution of the present invention is:

[0007] Microgrid optimization scheduling methods that consider system operation risks and user satisfaction include:

[0008] Step 1: Generate power forecast curves for wind power and photovoltaic new energy output within the microgrid, use the Monte Carlo simulation method to obtain the forecast error range to generate a typical scenario set, and obtain the wind and solar forecast error probability density distribution function through the data fitting method;

[0009] Step 2: Use the risk value and conditional risk value theory to quantify the increased microgrid operation risk caused by high or low output of wind power and photovoltaic power;

[0010] Step 3: Based on the time-of-use electricity price and incentive mechanism, a demand-side response model for transferable and interruptible loads is established to obtain the dispatching cost of adjustable loads in the microgrid system.

[0011] Step 4: Taking the minimum sum of the conditional risk value cost and the microgrid system operating cost as the objective function and the user satisfaction level under the demand-side response as the constraint, a microgrid optimization scheduling model considering the operating risk and user satisfaction is established;

[0012] Step 5: Use the genetic algorithm to solve the model and obtain the best results for optimizing the dispatch of the grid-connected microgrid system containing gas turbines, wind power, photovoltaics, energy storage and adjustable loads.

[0013] Methods for obtaining the probability density distribution function of wind and solar forecast errors include:

[0014] Step 1.1: During actual operation, the relationship between the wind turbine's power generation and wind speed is shown in the following formula:

[0015] Where, P w,t represents the output power of the wind turbine generator set at time t; v t represents the wind speed at time t; v in Indicates the cut-in wind speed; v out Indicates the cut-out wind speed; v e Indicates rated wind speed; P WN Indicates the rated output power of the wind turbine;

[0016] The relationship between the power generation power of a photovoltaic unit and the changes in solar radiation intensity and ambient temperature is shown in the following formula:

[0017] Where, P pv,t represents the output power of the photovoltaic generator set at time t; P stc is the output under standard conditions, I r,t is the actual solar radiation intensity at time t; I stc =1kW / m 2 , a T is the power temperature coefficient of the photovoltaic panel; T t is the temperature of the photovoltaic panel at time t, Tstc =25℃;

[0018] According to the forecast values ​​of wind speed, solar radiation intensity and temperature meteorological conditions, the average power forecast curve of wind power and photovoltaic power is generated, which is recorded as X = (X w ,X pv ), the prediction period is T, the sampling interval is Δt, X w and X pv are numerical vectors of 1×(T / Δt);

[0019] Step 1.2: Wind power and photovoltaic power generation output are random, volatile, and uncertain. The uncertainty of wind and photovoltaic prediction error is described by the following formula:

[0020] Where, Contribute to the forecast of wind and light, P k,t is the actual output of wind and solar power; ΔP k,t is the wind and solar prediction error, and The upper and lower limits of wind and solar power output prediction error; rut and rlt 0-1 variables indicating that the wind and solar power output is too high or too low;

[0021] The Monte Carlo simulation method is used to randomly obtain the wind and solar forecast error to generate m typical scene sets, which are denoted as Y = (Y w ,Y pv ), Y w and Y pv are numerical vectors of m×(T / Δt);

[0022] The probability density distribution function of wind and solar power forecast errors is calculated using mathematical modeling methods and data fitting technology. The skewed distribution is used to fit the power forecast error of wind power output; the standard normal distribution is used to fit the power forecast error of photovoltaic output.

[0023] Where x is the power prediction error of wind power output, μ w , σ w and λ w are the location parameter, scale parameter and skewness parameter of the skewed distribution respectively; y is the power prediction error of photovoltaic output, μ pv and σ pv are the mean and variance of the normal distribution respectively.

[0024] The risk value and conditional risk value theory are used to quantify the increased microgrid operation risks caused by high or low output of wind power and photovoltaic power, including:

[0025] Step 2.1: Use penalties to quantify the increased risk cost of over- or under-output caused by wind and solar power. Risk measurement is achieved using VaR (Value at Risk) and CvaR (Conditional Value at Risk). VaR is the maximum expected risk loss at a certain confidence level, assuming that the probability of risk loss at that confidence level will not exceed the VaR value.

[0026] For the loss function f(x,y), x and y are the optimization decision variables and the random variables that determine the system risk loss, respectively. The probability density function of y is ρ(y). Assuming α is the boundary value of f(x,y), the distribution function of the loss function f(x,y) is not greater than the boundary value α:

[0027] The VaR value under a given confidence level β∈(0,1) is α β (x) represents: VaR = α β (x)=min{α∈R;ψ(x,α)≥β};

[0028] CVaR reflects the portfolio risk by measuring the average loss that exceeds VaR, including quantile and tail risks; CVaR is defined as the conditional mean of the loss that exceeds VaR under the same confidence level β, expressed as φ β (x) represents the basic calculation method as follows:

[0029] To facilitate the solution, use the function F β (x,α) represents the CVaR value, that is:

[0030] Where α is the VaR value under the confidence level β; [f(x,y)-α] + represents max[0,f(x,y)-α];

[0031] Use the historical data of random variable y to estimate the above formula and discretize it:

[0032] Where, is the estimated value of CVaR, y k is the kth group of sample data of y, with a total of m groups;

[0033] Step 2.2: Use the CVaR theory to quantify the increased microgrid operation risk caused by excessively high or low wind and solar output. If the actual wind and solar output is excessively high, this is called underestimated risk, which can lead to oversupply and curtailment of wind and solar power.

[0034] Where, and They represent the risk costs of wind power and photovoltaic power abandonment at time t, α W and α PV are the critical values ​​of wind and solar curtailment risk losses, It represents the risk cost of renewable energy curtailment caused by the uncertainty of wind and solar power forecast errors; and Represent the penalty costs for wind and solar curtailment, and represent the penalty coefficients for wind and solar curtailment, and They represent the wind power curtailment and electricity power curtailment at time t under scenario m, respectively;

[0035] If the actual output of wind power and photovoltaic power is lower than expected, it is called overestimation risk. Overestimation risk may lead to supply shortage and load shedding:

[0036] Where, represents the risk cost of load shedding in the microgrid system, α LD is the critical value of load shedding risk loss; Penalty fees for load shedding, is the load shedding penalty coefficient, represents the load shedding power at time t under scenario m;

[0037] The total operating risk cost CVaR of the microgrid system can be expressed as U t +S t ≤1

[0038] Where U t and S t are 0-1 state variables indicating whether the microgrid system is at risk of power abandonment or load shedding at time t.

[0039] The demand-side response model for shiftable and interruptible loads based on time-of-use electricity prices and incentive mechanisms includes:

[0040] Step 3.1: The loads in the microgrid system mainly include rigid loads and flexible loads. Rigid loads cannot be regulated by the microgrid and are denoted as P. basi , flexible load is divided into transferable load P tran and interruptible load P disr Two categories, the total load P in the microgrid system load Expressed as: P load =P basi +Ptran +P disr ;

[0041] Establishing a transferable load demand response model based on time-of-use electricity price

[0042] Where, P tran,i and They represent the power before and after the transferable load demand response, T ij express

[0043] Transfer coefficient; λ tran represents the dispatch cost coefficient of load transfer, C tran Indicates transferable load

[0044] dispatch costs of demand response;

[0045] Establishing an interruptible load demand response model based on incentive policies

[0046] Where, P disr,i and They represent the power before and after the interruptible load demand response, ΔP disr,i represents the actual interruption power of the interruptible load at time i; disr represents the dispatch cost coefficient of load interruption, C disr represents the dispatch cost of interruptible load demand response;

[0047] Total load in the microgrid system under demand-side response and the total dispatch cost C DR Can be expressed as: C DR =C tran +C disr .

[0048] Establishing a microgrid optimization dispatch model that considers operational risks and user satisfaction includes:

[0049] Step 4.1: Take the minimum sum of the conditional risk value cost and the microgrid system operation cost as the objective function; min f = C OPE +C GRI +CVaR+C DR

[0050] Where C OPE is the fuel cost for gas turbine operation, P gas,t is the power generation capacity of the gas turbine, η gas is the power generation efficiency, Cgas is the market price of natural gas per cubic meter, H is the calorific value of natural gas per cubic meter, in kWh / m 3 ; C GRI is the cost of microgrid to interact with the main grid, λ buy,t and λ sell,t Represent the electricity purchase price and electricity sales price respectively, P buy,t and P sell,t Represent the purchased power and sold power respectively; CVaR represents the total operating risk cost of the microgrid system when considering the uncertainty of wind and solar power; C DR It represents the total dispatch cost of adjustable load in the microgrid system under demand-side response.

[0051] The constraints include system power balance constraints, unit output constraints, microgrid and main grid power interaction constraints, demand response constraints, and user satisfaction constraints;

[0052] Power balance constraints

[0053] Where, P g is the output of the gas turbine, and are the charge and discharge power of electrochemical energy storage, respectively;

[0054] Unit output constraints

[0055] Gas turbine output constraints

[0056] Renewable energy output constraints

[0057] Energy storage system output constraints E min ≤E ESS,t ≤E max E ESS,0 =E ESS,T

[0058] Where, is the maximum output power of the gas turbine; ΔP g,t is the change in gas turbine output power, are the upper and lower limits of the gas turbine ramp rate, respectively; are the maximum outputs of wind power and photovoltaic power in time period t respectively; They are the maximum charging power limit and the maximum discharging power limit of the energy storage system respectively; are 0-1 variables indicating whether the energy storage system is charging or discharging; E ESS,t is the amount of electricity stored in the energy storage system at time t, σ is the self-discharge rate of the battery device, η ch and η dis are charge and discharge efficiency respectively; E min and E max are the lower and upper limits of the amount of electricity that can be stored in the energy storage system; E ESS,0 Indicates the amount of energy stored at the initial moment of energy storage, E ESS,T Indicates the amount of electricity stored at the last moment of the energy storage cycle;

[0059] Interaction constraints between microgrid and main grid power

[0060] Where, The upper limits of the power that the IES can purchase from the grid and sell to the grid respectively; are 0-1 variables indicating whether the microgrid system is purchasing or selling electricity;

[0061] Demand response constraint 0≤-T ii ≤η max

[0062] Where, T ii and The numerical value indicates the actual transfer ratio and maximum transfer ratio of the load that can be transferred at time i. During the transfer process, the total load must remain unchanged. represents the maximum interruption power of the interruptible load at time i;

[0063] User satisfaction constraints

[0064] In the formula, the user satisfaction level ζ is introduced to measure the user's electricity experience, θ cu Numerical measurement of the worst acceptable user experience of electricity consumption;

[0065] By analyzing the changes in microgrid operation risk cost (CvaR) under different user satisfaction levels, the effectiveness of demand-side response in reducing power system operation risks is evaluated.

[0066] Methods for evaluating the effectiveness of demand-side response in reducing power system operational risks include:

[0067] Step 5.1. Population initialization: Use a random simulation method to generate a population of n individuals. Input the output forecast curves of the gas turbine, wind power, and photovoltaic power generation, the performance parameters of the wind power, photovoltaic power generation, energy storage, and tie-line equipment in the microgrid system, the load type, adjustable ratio, and time-of-use electricity price, and determine the boundary range of the population individuals.

[0068] Step 5.2: Population evaluation: Take the objective function as the fitness function, calculate the fitness of n individuals, and evaluate each individual based on the fitness to determine its quality.

[0069] Step 5.3: Population selection: Based on the fitness of individuals in the population, select individuals with high fitness from the current population through methods such as roulette. Roulette determines the number of genes that each individual inherits into the next generation with a probability proportional to fitness.

[0070] Step 5.4: Population crossover: Generate new individuals according to a certain crossover probability and crossover method;

[0071] Step 5.5: Population mutation: Generate new individuals according to a certain mutation probability and mutation method;

[0072] Step 5.6: Generate a new generation of population; generate a new generation of population based on population selection, crossover and mutation operations, and determine whether it meets the end conditions set by the number of iterations or accuracy. If so, output the best individual and its optimal solution and end. Otherwise, repeat steps 5.2 to 5.6.

[0073] Beneficial effects of the present invention:

[0074] Aiming at the risks of load shedding and power abandonment caused by the uncertainty of renewable energy prediction errors in the operation of microgrid systems, the present invention adopts the risk value theory VaR and the conditional risk value theory CVaR to quantify the risk cost, and incorporates the risk cost into the cost model of microgrid optimization scheduling; improving the system's flexibility and adjustment capability is an effective way to deal with the uncertainty risk of new energy predictions, considering the demand-side response can effectively reduce the operation risk of microgrid systems, incorporating the adjustable load scheduling cost into the microgrid economic model, and taking into account the user's electricity experience, a microgrid optimization scheduling method considering system operation risks and user satisfaction is proposed to achieve the characterization and control of microgrid system operation risks.

[0075] Advantages of the present invention:

[0076] The present invention takes into account the uncertainty of prediction errors of renewable energy sources such as wind and solar power, and adopts the CVaR theory to quantify the risk size of the uncertainty factors of renewable energy output, thereby establishing a risk cost function and avoiding the situation where the economic optimization scheduling results are too idealized.

[0077] The present invention fully exploits the load-side flexibility adjustment capability, introduces the adjustable load dispatching cost into the optimization objective function, and introduces user satisfaction into the constraint condition, while taking into account the impact of demand-side response on the microgrid operation economy, flexibility and user electricity experience.

[0078] The present invention guides users to participate in demand-side response based on time-of-use electricity prices and incentive mechanisms to reduce electricity costs and risk costs, effectively balances the operating costs and operating risks of microgrids, improves the risk management capabilities of microgrids, provides microgrids with more flexible optimization scheduling solutions, and enhances the safety, economy, stability and reliability of microgrids. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] FIG1 is a schematic diagram of the process of the present invention;

[0080] FIG2 is a schematic diagram of the solution process of the microgrid optimization scheduling model of the present invention. DETAILED DESCRIPTION

[0081] The microgrid optimization scheduling method considering system operation risk and user satisfaction includes the following steps:

[0082] Step 1: Generate power prediction curves for wind power and photovoltaic new energy output within the microgrid, use the Monte Carlo simulation method to obtain the prediction error range to generate a typical scenario set, and obtain the wind and solar prediction error probability density distribution function through the data fitting method.

[0083] Step 2: Use risk value and conditional risk value theory to quantify the increased microgrid operation risk caused by high or low output of wind power and photovoltaic power.

[0084] Step 3: Based on the time-of-use electricity price and incentive mechanism, a demand-side response model for transferable loads and interruptible loads is established to obtain the dispatching cost of the adjustable loads in the microgrid system.

[0085] Step 4: Taking the minimization of the sum of the conditional risk value cost and the microgrid system operating cost as the objective function and the user satisfaction level under the demand-side response as the constraint, a microgrid optimization scheduling model considering operation risk and user satisfaction is established.

[0086] Step 5: Use genetic algorithm to solve the model and obtain the best results for optimizing the dispatch of the grid-connected microgrid system containing gas turbines, wind power, photovoltaics, energy storage and adjustable loads.

[0087] In step 1, the day-ahead wind power and photovoltaic power forecast curves are generated based on meteorological conditions such as wind speed and solar radiation intensity. The Monte Carlo simulation method is used to obtain the forecast error range to generate a typical scenario set, and the wind and solar forecast error probability density distribution function is obtained through the data fitting method, which includes the following steps:

[0088] Step 1.1 During actual operation, the relationship between the wind turbine's power generation and wind speed is shown in the following formula:

[0089] Where, P w,t represents the output power of the wind turbine generator set at time t; v t represents the wind speed at time t; v in Indicates the cut-in wind speed; v out Indicates the cut-out wind speed; v e Indicates rated wind speed; P WN Indicates the rated output power of the wind turbine.

[0090] The relationship between the power generation power of a photovoltaic unit and the changes in solar radiation intensity and ambient temperature is shown in the following formula:

[0091] Where, P pv,t represents the output power of the photovoltaic generator set at time t; P stc is the output under standard conditions, I r,t is the actual solar radiation intensity at time t; I stc =1kW / m 2 , a T is the power temperature coefficient of the photovoltaic panel; T t is the temperature of the photovoltaic panel at time t, T stc =25℃.

[0092] According to the forecast values ​​of meteorological conditions such as wind speed, solar radiation intensity, temperature, etc., the average power forecast curve of wind power and photovoltaic power can be generated, which is recorded as X = (X w ,X pv ), the prediction period is T, the sampling interval is Δt, X w and X pv are numerical vectors of 1×(T / Δt) respectively.

[0093] Step 1.2 The output of wind power and photovoltaic new energy is random, volatile and uncertain. The uncertainty of wind and photovoltaic prediction error is described by the following formula:

[0094] Where, Contribute to the forecast of wind and light, P k,t is the actual output of wind and solar power; ΔP k,t is the wind and solar prediction error, and are the upper and lower limits of wind and solar power output prediction error; r ut and r lt 0-1 variables indicating that the wind and solar power output is too high or too low.

[0095] The Monte Carlo simulation method is used to randomly obtain the wind and solar forecast error to generate m typical scene sets, which are denoted as Y = (Y w ,Y pv ), Y w and Y pv are respectively numerical vectors of m×(T / Δt).

[0096] The probability density distribution function of wind and solar power prediction errors is calculated using mathematical modeling methods and data fitting technology. The power prediction of wind turbine output is nonlinear, and the distribution is multimodal, skewed, and heavy-tailed. The skewed distribution is used to fit the power prediction error of wind power output; the standard normal distribution is used to fit the power prediction error of photovoltaic output.

[0097] Where x is the power prediction error of wind power output, μ w , σ w and λ w are the location parameter, scale parameter and skewness parameter of the skewed distribution respectively; y is the power prediction error of photovoltaic output, μ pv and σ pv are the mean and variance of the normal distribution respectively.

[0098] In step 2, the risk value and conditional risk value theory are used to quantify the increased microgrid operation risk caused by high or low output of wind power and photovoltaic power, which includes the following steps:

[0099] Step 2.1 In microgrid systems, the uncertainty of renewable energy output power can cause fluctuations, posing certain risks to the stable operation of the grid. Therefore, penalties are used to quantify the increased risk cost caused by excessive or insufficient output from wind and solar power. Value at Risk (VaR) and Conditional Value at Risk (CVaR) are commonly used risk measurement methods. VaR is the maximum expected risk loss at a certain confidence level, assuming that the risk loss will not exceed the VaR value at that confidence level. The basic calculation method is as follows:

[0100] For the loss function f(x,y), x and y are the optimization decision variables and the random variables that determine the system risk loss, respectively. The probability density function of y is ρ(y). Assuming α is the boundary value of f(x,y), the distribution function of the loss function f(x,y) is not greater than the boundary value α:

[0101] The VaR value under a given confidence level β∈(0,1) is α β (x) represents: VaR = α β (x)=min{α∈R;ψ(x,α)≥β}

[0102] Compared to VaR, which only considers the risk information under the quantile and ignores the risk information at the tail of the quantile, CVaR can better reflect the portfolio risk by measuring the average loss of the part exceeding VaR, including the quantile and its tail risk. CVaR is defined as the conditional mean of the loss exceeding VaR under the same confidence level β, with φ β (x) represents the basic calculation method as follows:

[0103] To facilitate the solution, another relatively simple function F is often used β (x,α) represents the CVaR value, that is:

[0104] Where α is the VaR value under the confidence level β; [f(x,y)-α] + It means max[0,f(x,y)-α].

[0105] Usually, the probability density function ρ(y) is difficult to obtain. The historical data of the random variable y can be used to estimate the above formula and discretize it:

[0106] Where, is the estimated value of CVaR, y k is the kth group of sample data of y, with a total of m groups.

[0107] Step 2.2 uses the CVaR theory to quantify the increased microgrid operation risk caused by excessively high or low output of wind power and photovoltaic power. If the actual output of wind power and photovoltaic power is too high, it is called underestimated risk. Underestimation of risk may lead to oversupply and curtailment of wind and solar power:

[0108] Where, and They represent the risk costs of wind power and photovoltaic power abandonment at time t, α W and α PV are the critical values ​​of wind and solar curtailment risk losses, It represents the risk cost of renewable energy curtailment caused by the uncertainty of wind and solar power forecast errors; and Represent the penalty costs for wind and solar curtailment, and represent the penalty coefficients for wind and solar curtailment, and They represent the wind power curtailment and electricity power curtailment at time t under scenario m, respectively.

[0109] If the actual output of wind power and photovoltaic power is lower than expected, it is called overestimation risk. Overestimation risk may lead to supply shortage and load shedding:

[0110] Where, represents the risk cost of load shedding in the microgrid system, α LD is the critical value of load shedding risk loss; Penalty fees for load shedding, is the load shedding penalty coefficient, Indicates the load shedding power at time t in scenario m.

[0111] The total operating risk cost CVaR of the microgrid system can be expressed as U t +S t ≤1

[0112] Where U t and S t are 0-1 state variables indicating whether the microgrid system is at risk of power abandonment or load shedding at time t.

[0113] In step 3, a demand-side response model for shiftable and interruptible loads is established based on time-of-use electricity prices and incentive mechanisms, which includes the following steps:

[0114] Step 3.1 The loads in the microgrid system mainly include rigid loads and flexible loads. Rigid loads cannot be regulated by the microgrid and are denoted as P. basi , flexible load is divided into transferable load P tran , interruptible load P disr Two categories, the total load P in the microgrid system load It can be expressed as: P load =P basi +P tran +P disr

[0115] Establishing a transferable load demand response model based on time-of-use electricity price

[0116] Where, P tran,i and They represent the power before and after the transferable load demand response, T ij represents the transfer coefficient; λ tran represents the dispatch cost coefficient of load transfer, C tran represents the dispatch cost of shiftable load demand response.

[0117] Establishing an interruptible load demand response model based on incentive policies

[0118] Where, P disr,i and They represent the power before and after the interruptible load demand response, ΔP disr,i represents the actual interruption power of the interruptible load at time i; disr represents the dispatch cost coefficient of load interruption, C disr Represents the dispatch cost of interruptible load demand response.

[0119] Total load in the microgrid system under demand-side response and the total dispatch cost C DR Can be expressed as: C DR =C tran +C disr

[0120] In step 4, the minimum sum of the conditional risk value cost and the microgrid system operating cost is used as the objective function, and the user satisfaction level under the demand-side response is used as a constraint. A microgrid optimization scheduling model considering operation risk and user satisfaction is established, which includes the following steps:

[0121] Step 4.1 Take the minimum sum of the conditional risk value cost and the microgrid system operation cost as the objective function min f = C OPE +C GRI +CVaR+C DR

[0122] Where C OPE is the fuel cost for gas turbine operation, P gas,t is the power generation capacity of the gas turbine, η gas is the power generation efficiency, C gas is the market price of natural gas per cubic meter, H is the calorific value of natural gas per cubic meter, in kWh / m 3 ; C GRI is the cost of microgrid to interact with the main grid, λ buy,t and λ sell,t Represent the electricity purchase price and electricity sales price respectively, P buy,t and P sell,t Represent the purchased power and sold power respectively; CVaR represents the total operating risk cost of the microgrid system when considering the uncertainty of wind and solar power; C DR It represents the total dispatch cost of adjustable load in the microgrid system under demand-side response.

[0123] The constraints in step 4.2 mainly include system power balance constraints, unit output constraints, microgrid and main grid power interaction constraints, demand response constraints, user satisfaction constraints, etc.

[0124] ① Power balance constraints

[0125] Where, P g is the output of the gas turbine, and are the charging and discharging power of electrochemical energy storage, respectively.

[0126] ② Unit output constraints

[0127] Gas turbine output constraints

[0128] Renewable energy output constraints

[0129] Energy storage system output constraints E min ≤E ESS,t ≤E max E ESS,0 =E ESS,T

[0130] Where, is the maximum output power of the gas turbine; ΔP g,t is the change in gas turbine output power, are the upper and lower limits of the gas turbine ramp rate, respectively; are the maximum outputs of wind power and photovoltaic power in time period t respectively; They are the maximum charging power limit and the maximum discharging power limit of the energy storage system respectively; are 0-1 variables indicating whether the energy storage system is charging or discharging; E ESS,t is the amount of electricity stored in the energy storage system at time t, σ is the self-discharge rate of the battery device, η ch and η dis are charge and discharge efficiency respectively; E min and E max are the lower and upper limits of the amount of electricity that can be stored in the energy storage system; E ESS,0 Indicates the amount of energy stored at the initial moment of energy storage, E ESS,T Indicates the amount of electricity stored at the last moment of the energy storage cycle.

[0131] ③ Interaction constraints between microgrid and main grid power

[0132] Where, The upper limits of the power that the IES can purchase from the grid and sell to the grid respectively; They are 0-1 variables indicating whether the microgrid system is purchasing or selling electricity.

[0133] ④Demand response constraint 0≤-T ii ≤η max

[0134] Where, T ii and The numerical value indicates the actual transfer ratio and maximum transfer ratio of the load that can be transferred at time i. During the transfer process, the total load must remain unchanged. It represents the maximum interruption power of the load that can be interrupted at time i.

[0135] ⑤User satisfaction constraints

[0136] In the formula, the user satisfaction level ζ is introduced to measure the user's electricity experience, θ cu Numerically measures the worst acceptable user experience of electricity consumption.

[0137] In step 5, a genetic algorithm is used to solve the model and obtain the optimal dispatch results of the grid-connected microgrid system containing gas turbines, wind power, photovoltaics, energy storage, and adjustable loads. The changes in the microgrid operation risk cost (CvaR) under different user satisfaction levels are analyzed to evaluate the effectiveness of demand-side response in reducing power system operation risks. The following steps are included:

[0138] Step 1: Population initialization. A random simulation method is used to generate a population of n individuals. The output forecast curves of gas turbines, wind power, and photovoltaic power are input, as well as the performance parameters of wind power, photovoltaic power, energy storage, and tie lines in the microgrid system. The load category, adjustable ratio, and time-of-use electricity price are input, and the boundary range of the population individuals is determined.

[0139] Step 2: Population evaluation. Take the objective function as the fitness function, calculate the fitness of n individuals, and evaluate each individual based on the size of the fitness to determine its quality.

[0140] Step 3: Population selection. Based on the fitness of individuals in the population, individuals with high fitness are selected from the current population through methods such as roulette. Roulette determines the number of genes each individual inherits into the next generation using a probability proportional to fitness.

[0141] Step 4: Population crossover. Generate new individuals according to a certain crossover probability and crossover method.

[0142] Step 5: Population mutation. Generate new individuals according to a certain mutation probability and mutation method.

[0143] Step 6 generates a new generation of populations. This generation is generated through population selection, crossover, and mutation operations. The process then determines whether the termination conditions set by the number of iterations or accuracy are met. If so, the best individual and its optimal solution are output and the process ends. Otherwise, repeat steps 2 to 6.

[0144] Features of the present invention:

[0145] 1. Based on the value at risk (VaR) and the conditional value at risk (CVaR), the present invention establishes a microgrid risk assessment model that considers the uncertainty of wind and solar power forecast errors. The microgrid operation risk costs mainly include two types of risks: underestimated risk and overestimated risk: ① If the actual output of wind power and photovoltaic power is too high, it is called underestimated risk. Underestimated risk may lead to oversupply and wind and solar power abandonment; ② If the actual output of wind power and photovoltaic power is too low, it is called overestimated risk. Overestimated risk may lead to supply shortage and load shedding. The quantitative characterization of risks is achieved.

[0146] 2. On the one hand, the present invention takes into account the uncertainty of wind and solar power output and incorporates the operating risk cost of the microgrid system into the optimization objective function. On the other hand, it introduces demand-side response to improve system flexibility, incorporates adjustable load scheduling costs into the optimization objective function, and introduces user satisfaction into the constraint conditions; it fully considers the operating costs and operating risks of the microgrid system and realizes flexible risk management and control.

Claims

1. A microgrid optimal scheduling method considering system operation risks and user satisfaction, characterized in that: The method includes: Step 1: Generate the power prediction curves of wind power and photovoltaic new energy output in the microgrid, obtain the prediction error range by using the Monte Carlo simulation method to generate a typical scenario set, and obtain the wind-solar prediction error probability density distribution function through the data fitting method; Step 2: Use the value at risk and conditional value at risk theories to quantify the increased operation risks of the microgrid caused by the over-high or over-low output of wind power and photovoltaic; Step 3: Establish a demand response model for shiftable loads and interruptible loads based on time-of-use electricity prices and incentive mechanisms, and obtain the scheduling costs of adjustable loads in the microgrid system; Step 4: Take the sum of the conditional value at risk cost and the microgrid system operation cost as the objective function, and the user satisfaction level under demand response as the constraint, and establish a microgrid optimal scheduling model considering operation risks and user satisfaction; Step 5: Use the genetic algorithm to solve the model and obtain the optimal scheduling result of the grid-connected microgrid system including gas turbines, wind power, photovoltaic, energy storage, and adjustable loads.

2. The microgrid optimal scheduling method according to claim 1, considering system operation risks and user satisfaction, characterized in that: The method for obtaining the wind-solar prediction error probability density distribution function includes: Step 1.

1. During the actual operation, the relationship between the power generation of the wind turbine and the change in wind speed is shown by the following formula: Where, P w,t represents the output power of the wind turbine at time t; v t represents the wind speed magnitude at time t; v in represents the cut-in wind speed; v out represents the cut-out wind speed; v e represents the rated wind speed; P WN represents the rated output power of the wind turbine; The relationship between the power generation of a photovoltaic unit and the changes in solar radiation intensity and ambient temperature is shown by the following formula: Wherein, P pv,t represents the output power of the photovoltaic power generation unit at time t; P stc is the output under standard conditions, and I r,t is the actual solar radiation intensity at time t; I stc = 1kW / m 2 , a T is the power temperature coefficient of the photovoltaic panel; T t is the temperature of the photovoltaic panel at time t, and T stc = 25°C; According to the predicted values of the current wind speed, solar radiation intensity, and temperature meteorological conditions, generate the predicted curves of the average power of wind power and photovoltaic power, denoted as X = (X w , X pv ), the prediction period is T, the sampling interval is Δt, X w and X pv are numerical vectors of 1×(T / Δt) respectively; Step 1.

2. The output of wind power and photovoltaic new energy power generation is random, volatile, and uncertain. The uncertainty of the wind-solar prediction error is described by the following formula: In the formula, is the predicted output of wind and light, P k,t is the actual output of wind and light; ΔP k,t is the prediction error of wind and light, and are the upper and lower limit values of the prediction error of wind and light output; r ut and r lt respectively represent wind and light 0-1 variables for over-high or over-low output; Use the Monte Carlo simulation method to randomly obtain the wind and light prediction errors to generate m typical scenario sets, denoted as Y = (Y w , Y pv ), where Y w and Y pv are numerical vectors of m × (T / Δt) respectively; Adopt the mathematical modeling method and data fitting technology for the probability density distribution function of wind power and photovoltaic power prediction errors, use the skewed distribution to fit the power prediction error of wind power output, and use the standard normal distribution to fit the power prediction error of photovoltaic power output; where x is the power prediction error of wind power output, μ w , σ w and λ w are the location parameter, scale parameter and skewness parameter of this skewed distribution respectively; y is the power prediction error of photovoltaic output, μ pv and σ pv are the mean and variance of this normal distribution respectively.

3. The microgrid optimal scheduling method according to claim 1, considering system operation risks and user satisfaction, characterized in that: Using the value at risk and conditional value at risk theories to quantify the increased operation risks of the microgrid caused by the over-high or over-low output of wind power and photovoltaic includes: Step 2.1: Quantify the increased risk cost caused by the over-high or over-low output of wind-solar in the form of punishment. The risk measurement is realized by using value at risk VaR and conditional value at risk CvaR. VaR is the expected maximum risk loss at a certain confidence level, that is, it is considered that the risk loss will not exceed the VaR value with the probability of this confidence level; For the loss function f(x, y), where x and y are the optimization decision variable and the random variable determining the system risk loss respectively, and the probability density function of y is ρ(y). Assuming that α is the boundary value of f(x, y), the distribution function of the loss function f(x, y) not greater than the boundary value α is as follows: The VaR value at a given confidence level β ∈ (0, 1) is denoted by α β (x) indicates that: VaR = α β (x) = min{α ∈ R; ψ(x, α) ≥ β}; CVaR encompasses the quantile and its tail risk by measuring the average loss exceeding VaR, reflecting the portfolio risk; the definition of CVaR is the conditional mean of the loss exceeding VaR at the same confidence level β, denoted by φ β (x), and the basic calculation method is as follows: For the convenience of solving, the CVaR value is represented by the function F β (x, α), that is: Where α is the VaR value at confidence level β; [f(x, y) - α] + represents max[0, f(x, y) - α]; Estimate the above formula using the historical data of the random variable y and discretize it: In the formula, is the estimated value of CVaR, and y k is the k-th group of sample data of y, with a total of m groups; Step 2.2: Quantify the increased operation risk of the microgrid caused by the over- or under-generation of wind power and photovoltaic power using the CVaR theory. If the actual generation of wind power and photovoltaic power is higher, it is called the underestimated risk, which may lead to curtailment of wind and light due to oversupply. In the formula, And respectively represent the risk costs generated by wind power and PV curtailment at time t, α W and α PV are respectively the critical values of the losses of wind curtailment risk and PV curtailment risk, Indicates the risk cost of new energy curtailment conditions considering the uncertainty of wind and light prediction errors; and respectively represent the curtailment of wind and curtailment of solar penalty costs, And respectively represent the curtailment of wind and curtailment of solar penalty factors, And respectively represent the wind curtailment power and power curtailment power at time t in scenario m; If the actual output of wind power and photovoltaic power is on the low side, it is called overestimation risk. Overestimation risk may lead to power cuts due to supply falling short of demand: In the formula, Denote the risk cost generated by load shedding in the microgrid system, α LD is the critical value of the risk loss of load shedding; is the load shedding penalty cost, is the load shedding penalty factor, represents the load shedding power at time t in scenario m; The total operating risk cost CVaR of the microgrid system can be expressed as U t +S t ≤1 Where, U t and S t are 0-1 state variables representing whether the microgrid system is in the risk of curtailing electricity or load shedding at time t, respectively.

4. The microgrid optimal scheduling method according to claim 1, considering system operation risks and user satisfaction, characterized in that: Establishing a demand response model for shiftable loads and interruptible loads based on time-of-use electricity prices and incentive mechanisms includes: Step 3.

1. The loads in the microgrid system mainly include two categories: rigid loads and flexible loads. Rigid loads cannot accept the regulation of the microgrid and are denoted as P basi , and flexible loads are divided into shiftable loads P tran and interruptible loads P disr . The total load P load in the microgrid system is expressed as: P load = P basi + P tran + P disr ; Establish a Transferable Load Demand Response Model Based on Time-of-Use Electricity Price where P tran,i and respectively represent the power before and after the transferable load demand response, T ij represents the transfer coefficient; λ tran represents the scheduling cost coefficient of load transfer, C tran represents the scheduling cost of the transferable load demand response; Establish an interruptible load demand response model based on incentive policies Wherein, P disr,i and respectively represent the power before and after the interruptible load demand response, ΔP disr,i represents the actual interruption power of the interruptible load at time i; λ disr represents the scheduling cost coefficient of load interruption, C disr represents the scheduling cost of the interruptible load demand response; Total Load within the Microgrid System under Demand-Side Response and the total scheduling cost C DR can be respectively expressed as: C DR = C tran + C disr .

5. The microgrid optimal scheduling method according to claim 1, considering system operation risks and user satisfaction, characterized in that: Establishing a microgrid optimal scheduling model considering operation risks and user satisfaction includes: Step 4.1: Taking the minimum of the sum of the conditional value-at-risk cost and the microgrid system operation cost as the objective function; min f = C OPE + C GRI + CVaR + C DR Where, C OPE is the fuel cost of the gas turbine operation, P gas,t is the power generation capacity of the gas turbine, η gas is the power generation efficiency, C gas is the market price of natural gas per cubic meter, H is the calorific value of natural gas per cubic meter, with the unit of kWh / m 3 ; C GRI is the cost of power interaction between the microgrid and the main grid, λ buy,t and λ sell,t represent the power purchase price and the power selling price respectively, P buy,t and P sell,t represent the power purchase capacity and the power selling capacity respectively; CVaR represents the total operation risk cost of the microgrid system considering the uncertainty of wind and light; C DR represents the total scheduling cost of the adjustable load in the microgrid system under demand response.

6. The microgrid optimal scheduling method according to claim 5, considering system operation risks and user satisfaction, characterized in that: The constraint conditions include system power and energy balance constraints, unit output constraints, microgrid-main grid power interaction constraints, demand response constraints, and user satisfaction constraints; Power and electricity balance constraint Wherein, P g is the output of the gas turbine, and are respectively the charging and discharging electrical power of the electrochemical energy storage; Unit output constraint Gas turbine output constraint Renewable energy output constraint Output constraint of energy storage system E min ≤E ESS,t ≤E max E ESS,0 = E ESS,T In the formula, is the maximum output power of the gas turbine; ΔP g,t is the change in the output power of the gas turbine, They are the upper and lower limits of the gas turbine ramp rate, respectively; are the maximum outputs of wind power and photovoltaic power at time period t, respectively; They are the maximum charging power limit and the maximum discharging power limit of the energy storage system respectively; 0-1 variables indicating whether the energy storage system is charging or discharging, respectively; E ESS,t is the amount of electricity stored in the energy storage system at time t, σ is the self-discharge rate of the battery device, η ch and η dis are the charging and discharging efficiencies, respectively; E min and E max are the lower and upper limits of the allowable electricity storage of the energy storage system, respectively; E ESS,0 represents the stored power at the initial moment of energy storage, E ESS,T represents the stored power at the last moment of the energy storage cycle; Power Interaction Constraints between Microgrid and Main Grid In the formula, are the upper limits of the electric power purchased from the power grid and sold to the power grid by the IES, respectively; are respectively 0-1 variables indicating whether the microgrid system is purchasing or selling electricity; Demand response constraint 0 ≤ -T ii ≤ η max where T ii and Numerically represents the actual transfer ratio and the maximum transfer ratio of the transferable load at time i. During the transfer process, it is necessary to ensure that the total load remains unchanged; represents the maximum interruption power of the interruptible load at time i; User satisfaction constraint In the formula, the user satisfaction level ζ is introduced to measure the user's electricity consumption experience, and θ cu numerically measures the worst acceptable experience of the user's electricity consumption; 7. The microgrid optimal scheduling method considering system operation risk and user satisfaction according to claim 1, characterized in that: by analyzing the variation of the microgrid operation risk cost CvaR under different user satisfaction levels, the effectiveness of demand-side response in reducing the operation risk of the power system is evaluated.

8. The microgrid optimal scheduling method considering system operation risk and user satisfaction according to claim 7, characterized in that: the method for evaluating the effectiveness of demand-side response in reducing the operation risk of the power system includes: Step 5.1, population initialization; a population with an individual number of n is generated by using the random simulation method, the output prediction curves of gas turbines, wind power and photovoltaic power are input, the performance parameters of wind power, photovoltaic power, energy storage and tie line equipment in the microgrid system are input, the category, adjustable ratio and time-of-use electricity price of the load are input, and the boundary range of the population individuals is determined; Step 5.2, population evaluation; the objective function is used as the fitness function, the fitness of n individuals is calculated, and each individual is evaluated according to the size of the fitness to judge the quality; Step 5.3, population selection; according to the fitness size of the individuals in the population, the individuals with high fitness are selected from the current population by means of roulette wheel gambling, etc.; wherein the roulette wheel gambling determines the number of each individual inherited into the next generation population with a probability proportional to the fitness; Step 5.4, population crossover; new individuals are generated according to a certain crossover probability and crossover method; Step 5.5, population mutation; new individuals are generated according to a certain mutation probability and mutation method; Step 5.6, generating a new generation of population; a new generation of population is generated according to the population selection, crossover and mutation operations, and it is judged whether it meets the end conditions set by the iteration times or accuracy. If it meets, the best individual and its optimal solution are output and then ended. Otherwise, repeat steps 5.2 to 5.6.

Images