Cloud model-based current optimization method and system for AC / DC hybrid power distribution network containing PET (positron emission tomography)

Abstract

The invention discloses a day-ahead optimization method and system for a PET-containing AC / DC hybrid power distribution network based on a cloud model, and belongs to the field of day-ahead optimization scheduling of a power system. The method comprises the following steps: calculating three characteristic quantities of expectation, entropy and hyper-entropy of a cloud model to obtain the expectation, entropy and hyper-entropy of the cloud model; selecting a kernel function, and calculating probability density distribution of wind power and photovoltaic power prediction errors by adopting nonparametric kernel density estimation; setting confidence, finding the shortest interval to enable the probability that the prediction error value falls into the interval to be equal to the confidence, and obtaining upper and lower limits of wind power and photovoltaic power; constructing a day-ahead optimal scheduling model of the AC / DC hybrid power distribution network containing the PET considering the uncertainty of wind and light output prediction; and solving the model through a commercial solver to obtain a day-ahead optimal scheduling scheme of the AC / DC hybrid power distribution network containing the PET. According to the method, the uncertainty of wind and light output is considered, the power distribution network is allowed to accept more wind and light output, the total cost is reduced, and the power coordinating range of each port is expanded.

Description

Technical Field

[0001] The present invention belongs to the field of day-ahead optimization and dispatching of electric power systems, and relates to a method for taking into account the uncertainty of output of distributed power sources and research on day-ahead optimization and dispatching of distribution networks, and more specifically, to a day-ahead optimization method and system for a PET-containing AC / DC hybrid distribution network based on a cloud model. Background Art

[0002] With the development of power electronics technology, the AC / DC hybrid distribution network constructed based on power electronic transformers has made the distribution network present obvious zoning characteristics and stronger operation uncertainty.

[0003] Prior art document 1 (CN112615386B) discloses an optimization capacity determination method for an electric-thermal hybrid energy storage system for wind power consumption. Its shortcomings are that the wind power and photovoltaic power prediction errors have certain randomness and uncertainty, and due to the influence of various factors, the uncertainty distribution law of the output power level is not uniquely determined. Without considering the cloud model, it is impossible to qualitatively analyze the uncertainty of wind and solar power output prediction, and it is impossible to maintain the supply and demand balance of the system.

[0004] Prior art document 2 (CN110460036B) discloses a distributed optimization method for AC / DC distribution networks that takes into account wind power uncertainty. Its shortcomings are that for uncertainty analysis modeling, overall error modeling or single error modeling are difficult to show high reliability and adaptability at all times. Renewable energy sources such as wind power and photovoltaics connected to the AC / DC hybrid system have uncertainty, which will have an adverse impact on the operation of the distribution network. Summary of the invention

[0005] In order to solve the deficiencies in the prior art, the present invention provides a day-ahead optimization method for an AC / DC hybrid distribution network containing PET based on a cloud model, and studies the day-ahead optimization scheduling scheme for an AC / DC hybrid distribution network containing PET under the scenario of uncertainty in wind power and photovoltaic power prediction errors, thereby providing a reference for output planning of components such as micro gas turbines.

[0006] The present invention adopts the following technical solution.

[0007] The first aspect of the present invention provides a day-ahead optimization method for a PET-containing AC / DC hybrid distribution network based on a cloud model, comprising:

[0008] Step 1: Calculate the three characteristic quantities of the cloud model: expectation, entropy and super entropy;

[0009] Step 2: Use non-parametric kernel density estimation to solve the expectation, entropy and super entropy of the cloud model calculated in step 1 to obtain the probability density distribution of wind power and photovoltaic power prediction errors;

[0010] Step 3: Based on the probability density of wind power and photovoltaic power prediction errors obtained in step 2, set the confidence level, find the shortest interval so that the probability of the prediction error value falling within the interval is equal to the confidence level, and obtain the upper and lower limits of wind power and photovoltaic power;

[0011] Step 4: Based on the upper and lower limits of wind power and photovoltaic power obtained in step 3, a day-ahead optimal dispatch model of AC / DC hybrid distribution network including PET is constructed taking into account the uncertainty of wind and photovoltaic output forecast;

[0012] Step 5: Solve the day-ahead optimal dispatching model of the AC / DC hybrid distribution network containing PET taking into account the uncertainty of wind and solar power output forecast constructed in step 4, obtain the day-ahead optimal dispatching plan of the AC / DC hybrid distribution network containing PET, and realize the day-ahead optimization of the AC / DC hybrid distribution network containing PET based on the cloud model.

[0013] Preferably, in step 1, the three characteristic quantities of expectation, entropy and super entropy of the cloud model are expressed by the following formula:

[0014]

[0015] In the formula,

[0016] E P is the expected wind power and photovoltaic power prediction error,

[0017] E n is the entropy of wind power and photovoltaic power prediction error,

[0018] H e is the super entropy of wind power and photovoltaic power prediction error,

[0019] P i is the i-th sample value,

[0020] N is the total sample size,

[0021] S 2 is the sample variance of wind power and photovoltaic power prediction error.

[0022] Preferably, in step 2, non-parametric kernel density estimation is used to calculate the probability density f(P) of wind power and photovoltaic power prediction, which is expressed as the following formula:

[0023]

[0024] In the formula,

[0025] f(P) is the probability density of wind power and photovoltaic power prediction errors,

[0026] n is the interval sample number of wind power and photovoltaic power prediction errors,

[0027] P iis the i-th sample,

[0028] h is the bandwidth factor,

[0029] g is the kernel function, which is expressed as follows:

[0030]

[0031] In the formula,

[0032] μ j represents the expected wind power and photovoltaic power prediction error of the jth power station,

[0033] σ j It represents the standard deviation of the wind power and photovoltaic power prediction error of the j-th power station, which is solved by the entropy of the wind power and photovoltaic power prediction error of the j-th power station and the super entropy of the wind power and photovoltaic power prediction error of the j-th power station.

[0034] Preferably, in step 3, the wind power and photovoltaic power prediction error e is expressed by the following formula:

[0035] e=P fore -P true (4)

[0036] In the formula,

[0037] P fore is the predicted value of wind power and photovoltaic power,

[0038] P true is the actual value of wind power and photovoltaic power,

[0039] e is the wind power and photovoltaic power output prediction error value, and the distribution of e is represented by the probability density of wind power and photovoltaic power prediction errors.

[0040] Preferably, the wind power and photovoltaic output prediction error value e falls within the interval [e low ,e up ] the probability P(e low <e<e up ), expressed as follows:

[0041] P(e low <e<e up )=1-θ (5)

[0042] In the formula,

[0043] e low is the lower limit of the confidence interval,

[0044] e up is the upper limit of the confidence interval,

[0045] Interval [e low ,eup ] is the confidence interval at the confidence level of 1-θ.

[0046] Preferably, the interval [e low ,e up ] is the confidence interval under the confidence level of 1-θ, so the confidence interval of wind power and photovoltaic power prediction error is [P fore -e low ,P fore +e up ], the upper and lower limits of the confidence interval of wind power and photovoltaic power prediction errors are taken as the upper and lower limits of wind power and photovoltaic power.

[0047] Preferably, in step 4, a day-ahead optimal dispatching model of an AC / DC hybrid distribution network including PET taking into account the uncertainty of wind and solar power output forecast is constructed with the goal of minimizing the total cost, which is expressed by the following formula:

[0048] min f=F1+F2+F3 (6)

[0049] In the formula,

[0050] f is the total cost,

[0051] F1 is the cost of purchasing electricity from the main grid, expressed as follows:

[0052]

[0053] In the formula,

[0054] T is the total number of time periods in the optimization cycle,

[0055] λ M is the real-time electricity price of the main network at time t,

[0056] is the active power flowing from the main grid to the PET;

[0057] F2 is the cost of power generation by micro gas turbine, which is expressed as follows:

[0058]

[0059] In the formula,

[0060] λ MT is the cost coefficient of micro gas turbine power generation,

[0061] is the output power of the micro gas turbine at time t;

[0062] F3 is the charge and discharge loss cost of the energy storage device, expressed as follows:

[0063]

[0064] In the formula,

[0065] λ B is the loss cost coefficient of the energy storage device,

[0066] is the electrical energy stored in the energy storage device at time t.

[0067] Preferably, the constraints of the day-ahead optimal dispatch model of the AC / DC hybrid distribution network containing PET taking into account the uncertainty of wind and solar power output forecasting include power balance constraints of the AC / DC distribution network, operation constraints of the micro gas turbine and operation constraints of the energy storage device. The power balance constraints of the AC / DC distribution network are expressed by the following formula:

[0068]

[0069] In the formula,

[0070] is the active power flowing from the AC system to the PET,

[0071] is the active power flowing from the DC system to the PET,

[0072] is the active power flowing from the PET to the AC system,

[0073] is the active power flowing from the PET to the DC system,

[0074] is the output power of the wind turbine at time t,

[0075] is the load of the AC system at time t,

[0076] is the output power of photovoltaic power generation at time t,

[0077] is the output power of the photovoltaic wind power generation system at time t,

[0078] is the load of the DC system at time t;

[0079] The operating constraints of the micro gas turbine are expressed as follows:

[0080]

[0081] In the formula,

[0082] is the output power of the micro gas turbine at time t,

[0083] P MTm is the maximum output power of the micro gas turbine,

[0084] RU MT and RD MT They are respectively the upward and downward ramp rate limits of the micro gas turbine output power;

[0085] The operating constraints of the energy storage device are expressed as follows:

[0086]

[0087] In the formula,

[0088] is the charge capacity of the battery at time t,

[0089] P chm is the upper limit of the battery charge per hour,

[0090] is the discharge capacity of the battery at time t,

[0091] P dchm is the upper limit of battery discharge per hour,

[0092] is a binary variable, 1 means the energy storage device is in charging state at time t, 0 means it is not charging. is a binary variable, 1 means the energy storage device is in a discharging state at time t, and 0 means it is not discharging.

[0093] is the electrical energy stored in the energy storage device at time t,

[0094] B cap The upper limit of the amount of electrical energy that can be stored in the battery.

[0095] Preferably, the output power of the wind turbine at time t is The output power of photovoltaic power generation at time t The output power of the photovoltaic and wind power generation system at time t Both meet the upper and lower limits of wind power and photovoltaic power [P fore -e low ,P fore +e up ].

[0096] Embodiment 2 of the present invention provides a cloud model-based AC / DC hybrid distribution network day-ahead optimization system containing PET, which is used to run the cloud model-based AC / DC hybrid distribution network day-ahead optimization method containing PET, including:

[0097] Probability density solution module: used to calculate the three characteristic quantities of the cloud model, namely, expectation, entropy and hyperentropy, and use non-parametric kernel density estimation to solve the expectation, entropy and hyperentropy of the cloud model to obtain the probability density distribution of wind power and photovoltaic power prediction errors;

[0098] Power range solving module: used to set the confidence level according to the probability density of wind power and photovoltaic power prediction errors, find the shortest interval so that the probability of the prediction error value falling within the interval is equal to the confidence level, and obtain the upper and lower limits of wind power and photovoltaic power;

[0099] Modeling module: used to construct and obtain the day-ahead optimal dispatching model of AC / DC hybrid distribution network including PET taking into account the uncertainty of wind and solar power output forecast according to the upper and lower limits of wind power and photovoltaic power;

[0100] Model solving module: used to solve the day-ahead optimal dispatching model of the AC / DC hybrid distribution network containing PET taking into account the uncertainty of wind and solar power output forecast, obtain the day-ahead optimal dispatching plan of the AC / DC hybrid distribution network containing PET, and realize the day-ahead optimization of the AC / DC hybrid distribution network containing PET based on the cloud model.

[0101] The beneficial effects of the present invention are as follows:

[0102] The present invention proposes a day-ahead optimization method for a PET-containing AC / DC hybrid distribution network based on a cloud model. The cloud model and non-parametric kernel density estimation analyze the power prediction error from qualitative and quantitative perspectives, respectively, which helps to improve the expression accuracy of the prediction error distribution. The cloud model, non-parametric kernel density estimation and confidence interval are combined to obtain the upper and lower limits of wind power and photovoltaic power prediction errors to characterize the uncertainty of wind and solar power output prediction errors.

[0103] After considering the uncertainty of wind and solar power output, the control variables in the constraints, such as the output power of the micro gas turbine and the charge of the battery, are allowed to not meet the constraints. The AC / DC hybrid distribution network taking into account the uncertainty of wind and solar power output allows more wind and solar power output to be accepted, so the total cost of the system will be reduced. The output power of the AC / DC distribution network, the output power of the micro gas turbine, and the charging and discharging power of the energy storage device all involve the output values ​​of wind power and photovoltaic power, and are further constrained by the upper and lower limits of wind power and photovoltaic power. The range of power coordination of the micro gas turbine, energy storage device and each port of PET is expanded, and a more appropriate value can be selected under the premise of satisfying the objective function, that is, minimizing the cost;

[0104] By establishing a day-ahead optimal dispatch model for AC / DC hybrid distribution networks containing PET taking into account the uncertainty of wind and solar power output forecast, the problem of the uncertainty of renewable energy sources such as wind power and photovoltaics connected to AC / DC hybrid systems containing PET has been solved, which has an adverse impact on the operation of the distribution network. The effectiveness of the proposed model and method is verified through examples. The day-ahead optimal dispatch scheme for AC / DC hybrid distribution networks containing PET is studied in the scenario of uncertainty of wind power and photovoltaic power forecast errors, so as to provide a reference for the output plan of components such as micro gas turbines. BRIEF DESCRIPTION OF THE DRAWINGS

[0105] Figure 1 is a flow chart of the present invention,

[0106] Figure 2 Produce power diagrams for wind power forecasting,

[0107] Figure 3 For photovoltaic prediction output diagram,

[0108] Figure 4 This is a schematic diagram of wind power forecast output and its confidence interval.

[0109] Figure 5 This is a schematic diagram of photovoltaic predicted output and its confidence interval.

[0110] Figure 6 This is the schematic diagram of PET AC and DC port output power.

[0111] Figure 7 This is a schematic diagram of the arrangement of purchasing power from the main grid and the output of the micro gas turbine.

[0112] Figure 8 Schematic diagram of the charging and discharging arrangement of the energy storage device. DETAILED DESCRIPTION

[0113] In order to make the purpose, technical scheme and advantages of the present invention clearer, the technical scheme of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. The embodiments described in this application are only part of the embodiments of the present invention, not all of them. Based on the spirit of the present invention, other embodiments obtained by ordinary technicians in this field without creative work are all within the scope of protection of the present invention.

[0114] Example 1

[0115] Embodiment 1 of the present invention proposes a method for optimizing a hybrid AC / DC distribution network containing PET based on a cloud model, such as Figure 1 As shown, including:

[0116] Step 1: Apply the three characteristic quantities of the cloud model, namely expectation, entropy and hyperentropy, to express the uncertainty of wind power and photovoltaic power prediction, and obtain the expectation, entropy and hyperentropy of the cloud model;

[0117] The cloud model is an uncertainty conversion model between qualitative concepts and quantitative expressions based on probability statistics. Assume that U is a number space (one-dimensional, two-dimensional or multi-dimensional) represented by precise numerical values, and U corresponds to a qualitative concept For any element x in the number field U, there exists a stable tendency number It's called right The degree of certainty of the quantitative representation. The probability distribution of x belonging to U is called the cloud model.

[0118] Generally, the prediction errors of wind power and photovoltaic power have certain randomness and uncertainty, and due to the influence of various factors, the uncertainty distribution law of the output power meter level is not uniquely determined. The dual uncertainty impact brings difficulties to the grid dispatch based on wind and solar output. Therefore, the three characteristic quantities of the cloud model, namely expectation, entropy and super entropy, are used to qualitatively represent the uncertainty of wind power and photovoltaic power prediction. Expectation reflects the average stable tendency number in the entire cloud model. Entropy is a measure of uncertainty. In the cloud model, it reflects the degree of confusion of information. For a stable distribution, a smaller entropy means a higher certainty, while a larger entropy means a more evenly dispersed uncertainty. Super entropy can capture higher-order uncertainty characteristics. Compared with entropy, super entropy considers the probability distribution of stable tendency numbers rather than a simple probability density function, which helps to deeply understand the uncertainty in complex systems.

[0119] The three characteristic quantities of the cloud model, namely expectation, entropy and super entropy, are used to qualitatively represent the uncertainty of wind power and photovoltaic power forecasting, which can be expressed as follows:

[0120]

[0121] In the formula,

[0122] E P represents the expectation of wind power and photovoltaic power forecast error;

[0123] E n represents the entropy of wind power and photovoltaic power prediction error;

[0124] H e represents the hyperentropy of wind power and photovoltaic power prediction errors;

[0125] P i is the i-th sample value,

[0126] N is the total sample size,

[0127] S 2 is the sample variance of wind power and photovoltaic power prediction error.

[0128] Step 2: Select a kernel function and use non-parametric kernel density estimation to calculate the probability density distribution of wind power and photovoltaic power prediction errors based on the expectation, entropy and hyperentropy of the cloud model obtained in step 1 to obtain the probability density of wind power and photovoltaic power prediction errors;

[0129] Since cloud models are mainly used for qualitative analysis of the uncertainty of wind and solar power output forecast errors, and it is difficult to quantify them, nonparametric kernel density estimation and confidence interval methods are further used to quantify the uncertainty of wind and solar power output forecast errors.

[0130] The non-parametric kernel density estimation is used to quantitatively represent the uncertainty of wind power and photovoltaic power forecast errors. The non-parametric kernel density estimation of the probability density f(P) of wind power and photovoltaic power forecast errors is expressed as follows:

[0131]

[0132] In the formula,

[0133] n represents the number of samples in the wind power and photovoltaic power prediction error interval;

[0134] P i represents the i-th sample;

[0135] h represents the bandwidth factor;

[0136] g represents the kernel function, which is expressed as follows:

[0137]

[0138] In the formula,

[0139] μ j represents the expected wind power and photovoltaic power prediction error of the jth power station,

[0140] σ j It represents the standard deviation of the wind power and photovoltaic power prediction error of the j-th power station, which is obtained by solving the entropy of the wind power and photovoltaic power prediction error of the j-th power station and the super entropy of the wind power and photovoltaic power prediction error of the j-th power station. Refer to Formula 1. The standard deviation of the wind power and photovoltaic power prediction error of the j-th power station

[0141] Step 3: Based on the probability density of wind power and photovoltaic power prediction errors obtained in step 2, set the confidence level, and find the shortest interval so that the probability of the prediction error value falling within the interval is equal to the confidence level, and obtain the upper and lower limits of the wind power and photovoltaic power prediction errors.

[0142] e=P fore -P true (4)

[0143] In the formula,

[0144] P fore is the predicted value of wind power and photovoltaic power;

[0145] P true is the actual value of wind power and photovoltaic power;

[0146] e is the wind power and photovoltaic power output prediction error value, and the distribution of e is represented by the probability density of wind power and photovoltaic power prediction errors.

[0147] P(e low <e<e up )=1-θ (5)

[0148] In the formula,

[0149] Interval [e low ,e up ] is the confidence interval at the confidence level of 1-θ,

[0150] e low is the lower limit of the confidence interval,

[0151] e up is the upper limit of the confidence interval;

[0152] P(e low <e<e up ) is the wind power and photovoltaic output prediction error value e falling into the interval [e low ,e up ] is the probability within .

[0153] Interval [e low ,e up ] is the confidence interval under the confidence level of 1-θ, so the confidence interval of wind power and photovoltaic power prediction error is [P fore -e low ,P fore +e up ], the upper and lower limits of the confidence interval of wind power and photovoltaic power prediction errors are taken as the upper and lower limits of wind power and photovoltaic power.

[0154] Confidence intervals for wind and photovoltaic power forecasts are usually constructed by analyzing the sample statistics of their prediction errors and thereby estimating the fluctuation range of the overall parameters.

[0155] Step 4: Based on the upper and lower limits of wind power and photovoltaic power obtained in step 3, a day-ahead optimal dispatch model of AC / DC hybrid distribution network including PET is constructed taking into account the uncertainty of wind and photovoltaic output forecast;

[0156] An optimization model with the total system cost as the objective function, including the power exchange cost with the main grid, the power generation cost of the micro gas turbine and the charging and discharging loss cost of the energy storage device, was established. The model took into account constraints such as the power balance of the AC and DC distribution networks, the operation of the micro gas turbine and the operation of the energy storage device.

[0157] An optimization scheme is established that comprehensively considers the cost of purchasing electricity from the main grid, the cost of generating electricity from micro gas turbines, and the cost of charging and discharging loss of energy storage devices. A day-ahead optimization dispatch model for AC / DC hybrid distribution networks including PET is constructed with the goal of minimizing the total cost and taking into account the uncertainty of wind and solar power output forecasting:

[0158] min f=F1+F2+F3 (6)

[0159] In the formula,

[0160] f is the total cost,

[0161] F1 is the cost of purchasing electricity from the main grid,

[0162] F2 is the power generation cost of micro gas turbine,

[0163] F3 is the charging and discharging loss cost of the energy storage device.

[0164] The specific analysis of F1, F2 and F3 is as follows:

[0165] 1) Cost of purchasing electricity from the main grid

[0166] In addition to the AC distribution network and the DC distribution network, the AC / DC hybrid system is also connected to the main grid via PET. Assuming that the power exchange with the main grid is minimal, when there is a power shortage in the system itself, it is possible to consider purchasing electricity from the main grid to balance the power shortage of the AC / DC hybrid system. The objective function of the cost of purchasing electricity from the main grid is:

[0167]

[0168] In the formula,

[0169] T is the total number of time periods in the optimization cycle,

[0170] λ M is the real-time electricity price of the main grid at time t;

[0171] It is the active power flowing from the main grid to the PET.

[0172] 2) Micro gas turbine power generation cost

[0173] The controllable distributed power sources in the AC distribution network mainly include micro gas turbines, and their power generation cost can be approximately expressed by a linear function. The objective function of the power generation cost of micro gas turbines is:

[0174]

[0175] In the formula,

[0176] λ MT is the cost coefficient of power generation by micro gas turbine;

[0177] is the output power of the micro gas turbine at time t.

[0178] 3) Charging and discharging loss cost of energy storage device

[0179] The energy storage device will always be in a charging or discharging state during the operation of the system, and inevitable power loss will occur in this process. This loss is related to the power of the energy storage device charging and discharging. The objective function of the energy storage device charging and discharging loss cost is:

[0180]

[0181] In the formula,

[0182] λ B is the loss cost coefficient of the energy storage device;

[0183] is the electrical energy stored in the energy storage device at time t.

[0184] The constraints are analyzed as follows:

[0185] 1) Power balance constraints of AC and DC distribution networks

[0186]

[0187]

[0188] In the formula,

[0189] is the active power flowing from the AC system into the PET;

[0190] is the active power flowing from the DC system into the PET;

[0191] is the active power flowing from the PET to the AC system;

[0192] is the active power flowing from the PET to the DC system;

[0193] is the output power of the wind turbine at time t;

[0194] is the load of the AC system at time t;

[0195] is the output power of photovoltaic power generation at time t;

[0196] is the output power of the photovoltaic and wind power generation system at time t;

[0197] is the load of the DC system at time t.

[0198] Output power of wind turbine at time t The output power of photovoltaic power generation at time t The output power of the photovoltaic and wind power generation system at time t Both meet the upper and lower limits of wind power and photovoltaic power [P fore -e low ,P fore +e up ].

[0199] 2) Micro gas turbine operation constraints

[0200] The output power of the micro gas turbine is not large, and its output is generally controlled by the speed control system and the excitation system. Among them, the operation control of the speed control system is determined by the load level to adjust the active output of the gas turbine. In the optimization process, the output power of the micro gas turbine can be used as the control variable of the model, and its power generation constraints are as follows:

[0201]

[0202] In the formula,

[0203] is the output power of the micro gas turbine at time t;

[0204] P MTm is the maximum output power of the micro gas turbine;

[0205] RU MT and RD MT They are the upward and downward ramp rate limits of the micro gas turbine output power respectively.

[0206] 3) Energy storage device operation constraints

[0207] For the operation constraints of the energy storage system, the following four aspects are considered:

[0208] ① The charging and discharging power of the energy storage system at any time has upper and lower limits;

[0209] ② The energy storage system cannot be charged and discharged at the same time at any one time;

[0210] ③ The electric energy stored by the energy storage system at time t does not exceed the upper limit of the electric energy stored by the battery;

[0211] ④ The energy storage system will self-discharge at every moment, so the storage energy of the energy storage system at the next moment is equal to the sum of the remaining energy after the energy storage system self-discharge at the previous moment and the charging and discharging energy at this moment.

[0212] The four constraints can be expressed as follows:

[0213]

[0214] In the formula,

[0215] is the charge capacity of the battery at time t;

[0216] P chm The upper limit of the battery charge per hour;

[0217] is the discharge capacity of the battery at time t;

[0218] P dchm The upper limit of battery discharge per hour;

[0219] It is a binary variable, 1 means the energy storage device is in charging state at time t, and 0 means it is not charging;

[0220] It is a binary variable, 1 means the energy storage device is in the discharge state at time t, and 0 means it is not discharged;

[0221] is the electrical energy stored in the energy storage device at time t;

[0222] B cap The upper limit of the amount of electrical energy that can be stored in the battery.

[0223] The output power of the AC / DC distribution network, the output power of the micro gas turbine, and the charging and discharging power of the energy storage device all have their own operating constraints. In addition, the power balance constraints of the AC / DC distribution network must be met, which will involve the output values ​​of wind power and photovoltaics and will be affected by the upper and lower limits of wind power and photovoltaic power.

[0224] Step 5: Solve the day-ahead optimal dispatching model of the AC / DC hybrid distribution network containing PET in step 4 by using a commercial solver, obtain the day-ahead optimal dispatching scheme of the AC / DC hybrid distribution network containing PET, and realize the power flow exchange of the AC / DC hybrid distribution network containing PET based on the random error model of power output.

[0225] The beneficial effects of the present invention are as follows:

[0226] The present invention proposes a day-ahead optimization method for a PET-containing AC / DC hybrid distribution network based on a cloud model. The cloud model and non-parametric kernel density estimation analyze the power prediction error from qualitative and quantitative perspectives, respectively, which helps to improve the expression accuracy of the prediction error distribution. The cloud model, non-parametric kernel density estimation and confidence interval are combined to obtain the upper and lower limits of wind power and photovoltaic power prediction errors to characterize the uncertainty of wind and solar power output prediction errors.

[0227] After considering the uncertainty of wind and solar power output, the control variables in the constraints, such as the output power of the micro gas turbine and the charge of the battery, are allowed to not meet the constraints. The AC / DC hybrid distribution network taking into account the uncertainty of wind and solar power output allows more wind and solar power output to be accepted, so the total cost of the system will be reduced. The output power of the AC / DC distribution network, the output power of the micro gas turbine, and the charging and discharging power of the energy storage device all involve the output values ​​of wind power and photovoltaic power, and are further constrained by the upper and lower limits of wind power and photovoltaic power. The range of power coordination of the micro gas turbine, energy storage device and each port of PET is expanded, and a more appropriate value can be selected under the premise of satisfying the objective function, that is, minimizing the cost;

[0228] By establishing a day-ahead optimal dispatch model for AC / DC hybrid distribution networks containing PET taking into account the uncertainty of wind and solar power output forecast, the problem of the uncertainty of renewable energy sources such as wind power and photovoltaics connected to AC / DC hybrid systems containing PET has been solved, which has an adverse impact on the operation of the distribution network. The effectiveness of the proposed model and method is verified through examples. The day-ahead optimal dispatch scheme for AC / DC hybrid distribution networks containing PET is studied in the scenario of uncertainty of wind power and photovoltaic power forecast errors, so as to provide a reference for the output plan of components such as micro gas turbines.

[0229] Example 2

[0230] Embodiment 2 of the present invention proposes a cloud model-based AC / DC hybrid distribution network day-ahead optimization system containing PET, which is used to run a cloud model-based AC / DC hybrid distribution network day-ahead optimization method containing PET in Embodiment 1, including:

[0231] Probability density solution module: used to calculate the three characteristic quantities of the cloud model, namely, expectation, entropy and hyperentropy, and use non-parametric kernel density estimation to solve the expectation, entropy and hyperentropy of the cloud model to obtain the probability density distribution of wind power and photovoltaic power prediction errors;

[0232] Power range solving module: used to set the confidence level according to the probability density of wind power and photovoltaic power prediction errors, find the shortest interval so that the probability of the prediction error value falling within the interval is equal to the confidence level, and obtain the upper and lower limits of wind power and photovoltaic power;

[0233] Modeling module: used to construct and obtain the day-ahead optimal dispatching model of AC / DC hybrid distribution network including PET taking into account the uncertainty of wind and solar power output forecast according to the upper and lower limits of wind power and photovoltaic power;

[0234] Model solving module: used to solve the day-ahead optimal dispatching model of the AC / DC hybrid distribution network containing PET taking into account the uncertainty of wind and solar power output forecast, obtain the day-ahead optimal dispatching plan of the AC / DC hybrid distribution network containing PET, and realize the day-ahead optimization of the AC / DC hybrid distribution network containing PET based on the cloud model.

[0235] Example 3

[0236] This embodiment implements the method of the present invention through specific data. The results show that the present invention establishes a day-ahead optimization scheduling plan under two different scenarios by considering the uncertainty of distributed power output, providing a reference for the power output arrangement of micro gas turbines, energy storage devices and PET ports. The data of this embodiment are as follows:

[0237] The example test system used in the present invention is an AC / DC hybrid power distribution network containing PET, which includes a 10kV AC distribution network and a 750V DC distribution network, which are coupled and connected through a PET. The AC power grid includes wind turbines, micro gas turbines and AC loads, and the DC power grid includes photovoltaic power generation devices, energy storage devices, photovoltaic wind power generation systems and DC loads. The 24h data of AC loads, DC loads and real-time electricity prices of the main network are shown in Table 1.

[0238] Table 1 AC / DC load and main grid real-time electricity price data

[0239]

[0240]

[0241] The specific scenario is as follows:

[0242] Scenario 1: Optimal dispatch of AC / DC hybrid distribution network with PET without considering the uncertainty of wind power and photovoltaic power forecast error. The wind power and photovoltaic power forecast output curve is as follows: Figure 2 and Figure 3 shown.

[0243] Scenario 2: Set the confidence level to 95% and optimize the scheduling of AC / DC hybrid distribution network with PET taking into account the uncertainty of wind power and photovoltaic power forecast errors. The confidence intervals of wind power and photovoltaic output are as follows: Figure 4 and Figure 5 shown.

[0244] This paper uses a commercial solver to solve the problem, and finally obtains a comparison of the total network cost of the day-ahead optimal scheduling of the AC / DC hybrid distribution network under two different scenarios, as shown in Table 2.

[0245] As can be seen from Table 2, the total cost of taking into account the uncertainty of wind and solar output forecast errors is lower than the total cost without taking into account. When the confidence level is 95%, the total cost is 6.48% lower. The reason for this result is that after considering the uncertainty, the network is allowed to fail to meet the system safety operation constraints under certain probability conditions. The AC / DC hybrid distribution network that takes into account the uncertainty of wind and solar output forecast errors allows more wind and solar output to be accepted, so the total cost of the system will be reduced.

[0246] Table 2 Comparison of day-ahead optimization dispatch results of AC / DC hybrid distribution network

[0247]

[0248] Taking scenario 2 as an example, Figure 6 The output power diagram of the PET AC port and tributary port within 24 hours is given. A positive value indicates that power flows out of the AC / DC system, and a negative value indicates that power is injected into the AC / DC system. Figure 6 It can be seen that the port connected to the DC system needs to transmit power from the AC system through PET to supply the DC load during 1-8h and 17-24h due to insufficient photovoltaic output; and between 9-16h, because the photovoltaic output exceeds the load demand, the excess power is transmitted to the DC system after deducting the charging capacity of the energy storage device.

[0249] Figure 7 The results of power injection at the main grid and the dispatch output of the micro gas turbine are given. When the power price of the main grid is cheaper than the self-generated power of the AC / DC system, the system will purchase power from the main grid. Figure 7 It can be seen that there is power purchase behavior from the main grid at 3-6h and 19-22h. This is because the power purchase price of the main grid at this time is the off-peak price, which is significantly lower than the unit power generation cost of the micro gas turbine, and the power purchase of the main grid has also reached the upper limit, so the output of the micro gas turbine should be reduced as much as possible to reduce the system operation cost.

[0250] Figure 8 The dispatching output of the energy storage system is given. During 9-13h, due to the continuous increase in photovoltaic output, there is a surplus of power on the DC side and the load level is relatively low, so the energy storage system operates in a charging state. During 14-16h, when the power on the DC side is in short supply, the energy storage system releases electric energy to provide active power for nearby loads. Through the charging and discharging of the battery at different time periods, the supply and demand balance and economy of the system are achieved, and the safety of the system is effectively improved.

[0251] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A day-ahead optimization method for a PET-containing AC / DC hybrid distribution network based on a cloud model, characterized in that: Step 1: Calculate the three characteristic quantities of the cloud model: expectation, entropy and super entropy; Step 2: Use non-parametric kernel density estimation to solve the expectation, entropy and super entropy of the cloud model calculated in step 1 to obtain the probability density distribution of wind power and photovoltaic power prediction errors; Step 3: Based on the probability density of wind power and photovoltaic power prediction errors obtained in step 2, set the confidence level, find the shortest interval so that the probability of the prediction error value falling within the interval is equal to the confidence level, and obtain the upper and lower limits of wind power and photovoltaic power; Step 4: Based on the upper and lower limits of wind power and photovoltaic power obtained in step 3, a day-ahead optimal dispatch model of AC / DC hybrid distribution network including PET is constructed taking into account the uncertainty of wind and photovoltaic output forecast; Step 5: Solve the day-ahead optimal dispatching model of the AC / DC hybrid distribution network containing PET taking into account the uncertainty of wind and solar power output forecast constructed in step 4, obtain the day-ahead optimal dispatching plan of the AC / DC hybrid distribution network containing PET, and realize the day-ahead optimization of the AC / DC hybrid distribution network containing PET based on the cloud model.

2. The method for optimizing the day-ahead AC / DC hybrid distribution network containing PET based on a cloud model according to claim 1, characterized in that: In step 1, the three characteristic quantities of the cloud model, namely expectation, entropy and super entropy, are expressed as follows: In the formula, E P is the expected wind power and photovoltaic power prediction error, E n is the entropy of wind power and photovoltaic power prediction error, H e is the super entropy of wind power and photovoltaic power prediction error, P i is the i-th sample value, N is the total sample size, S 2 is the sample variance of wind power and photovoltaic power prediction error.

3. The method for optimizing the day-ahead AC / DC hybrid distribution network containing PET based on a cloud model according to claim 1, characterized in that: In step 2, non-parametric kernel density estimation is used to calculate the probability density f(P) of wind power and photovoltaic power forecast, which is expressed as follows: In the formula, f(P) is the probability density of wind power and photovoltaic power prediction errors, n is the interval sample number of wind power and photovoltaic power prediction errors, P i is the i-th sample, h is the bandwidth factor, g is the kernel function, which is expressed as follows: In the formula, μ j represents the expected wind power and photovoltaic power prediction error of the jth power station, σ j It represents the standard deviation of the wind power and photovoltaic power prediction error of the j-th power station, which is solved by the entropy of the wind power and photovoltaic power prediction error of the j-th power station and the super entropy of the wind power and photovoltaic power prediction error of the j-th power station.

4. The method for optimizing the day-ahead AC / DC hybrid distribution network containing PET based on a cloud model according to claim 1, characterized in that: In step 3, the wind power and photovoltaic power prediction error e is expressed as follows: e=P fore -P true (4) In the formula, P fore is the predicted value of wind power and photovoltaic power, P true is the actual value of wind power and photovoltaic power, e is the wind power and photovoltaic power output prediction error value, and the distribution of e is represented by the probability density of wind power and photovoltaic power prediction errors.

5. The method for optimizing the day-ahead AC / DC hybrid distribution network containing PET based on a cloud model according to claim 4, characterized in that: The wind power and photovoltaic output prediction error value e falls within the interval [e low ,e up ] the probability P(e low <e<e up ), expressed as follows: P(e low <e<e up )=1-θ (5) In the formula, e low is the lower limit of the confidence interval, e up is the upper limit of the confidence interval, Interval [e low ,e up ] is the confidence interval at the confidence level of 1-θ.

6. The method for optimizing the day-ahead AC / DC hybrid distribution network containing PET based on a cloud model according to claim 5, characterized in that: Interval [e low ,e up ] is the confidence interval under the confidence level of 1-θ, so the confidence interval of wind power and photovoltaic power prediction error is [P fore -e low ,P fore +e up ], the upper and lower limits of the confidence interval of wind power and photovoltaic power prediction errors are taken as the upper and lower limits of wind power and photovoltaic power.

7. The method for optimizing the day-ahead AC / DC hybrid distribution network containing PET based on a cloud model according to claim 1, characterized in that: In step 4, the day-ahead optimal dispatch model of the AC / DC hybrid distribution network including PET taking into account the uncertainty of wind and solar power output forecast is constructed with the goal of minimizing the total cost, which is expressed as follows: min f=F1+F2+F3 (6) In the formula, f is the total cost, F1 is the cost of purchasing electricity from the main grid, expressed as follows: In the formula, T is the total number of optimization periods. λ M is the real-time electricity price of the main network at time t, is the active power flowing from the main grid to the PET; F2 is the cost of power generation by micro gas turbine, which is expressed as follows: In the formula, λ MT is the cost coefficient of micro gas turbine power generation, is the output power of the micro gas turbine at time t; F3 is the charge and discharge loss cost of the energy storage device, expressed as follows: In the formula, λ B is the loss cost coefficient of the energy storage device, is the electrical energy stored in the energy storage device at time t.

8. The method for optimizing the day-ahead AC / DC hybrid distribution network containing PET based on a cloud model according to claim 7, characterized in that: The constraints of the day-ahead optimal dispatch model of the AC / DC hybrid distribution network with PET considering the uncertainty of wind and solar power output forecast include the power balance constraint of the AC / DC distribution network, the operation constraint of the micro gas turbine and the operation constraint of the energy storage device. The power balance constraint of the AC / DC distribution network is expressed as follows: In the formula, is the active power flowing from the AC system to the PET, is the active power flowing from the DC system to the PET, is the active power flowing from the PET to the AC system, is the active power flowing from the PET to the DC system, is the output power of the wind turbine at time t, is the load of the AC system at time t, is the output power of photovoltaic power generation at time t, is the output power of the photovoltaic wind power generation system at time t, is the load of the DC system at time t; The operating constraints of the micro gas turbine are expressed as follows: In the formula, is the output power of the micro gas turbine at time t, P MTm is the maximum output power of the micro gas turbine, RU MT and RD MT They are respectively the upward and downward ramp rate limits of the micro gas turbine output power and the operation constraints of the energy storage device, which are expressed by the following formula: In the formula, is the battery charge at time t, P chm The upper limit of the battery charge per hour. is the discharge capacity of the battery at time t, P dchm is the upper limit of battery discharge per hour, is a binary variable, 1 means the energy storage device is in charging state at time t, 0 means it is not charging. is a binary variable, 1 means the energy storage device is in a discharging state at time t, and 0 means it is not discharging. is the electrical energy stored in the energy storage device at time t, B cap The upper limit of the amount of electrical energy that can be stored in the battery.

9. The method for optimizing the day-ahead AC / DC hybrid distribution network containing PET based on a cloud model according to claim 8, characterized in that: Output power of wind turbine at time t The output power of photovoltaic power generation at time t The output power of the photovoltaic and wind power generation system at time t Both meet the upper and lower limits of wind power and photovoltaic power [P fore -e low ,P fore +e up ].

10. A cloud model-based AC / DC hybrid distribution network day-ahead optimization system containing PET, used to run a cloud model-based AC / DC hybrid distribution network day-ahead optimization method containing PET according to any one of claims 1 to 9, characterized in that: Probability density solution module: used to calculate the three characteristic quantities of the cloud model, namely, expectation, entropy and hyperentropy, and use non-parametric kernel density estimation to solve the expectation, entropy and hyperentropy of the cloud model to obtain the probability density distribution of wind power and photovoltaic power prediction errors; Power range solving module: used to set the confidence level according to the probability density of wind power and photovoltaic power prediction errors, find the shortest interval so that the probability of the prediction error value falling within the interval is equal to the confidence level, and obtain the upper and lower limits of wind power and photovoltaic power; Modeling module: used to construct and obtain the day-ahead optimal dispatching model of AC / DC hybrid distribution network including PET taking into account the uncertainty of wind and solar power output forecast according to the upper and lower limits of wind power and photovoltaic power; Model solving module: used to solve the day-ahead optimal dispatching model of the AC / DC hybrid distribution network containing PET taking into account the uncertainty of wind and solar power output forecast, obtain the day-ahead optimal dispatching plan of the AC / DC hybrid distribution network containing PET, and realize the day-ahead optimization of the AC / DC hybrid distribution network containing PET based on the cloud model.

Images