Power distribution network active and reactive coordination optimization control method based on distributed new energy

Abstract

The invention provides a power distribution network active and reactive coordinated optimization control method based on distributed new energy, relates to the technical field of voltage control of a power distribution network, and introduces a multi-scene random modeling method to reduce the influence of load and new energy output volatility and uncertain factors and ensure the robustness of an optimization control strategy. And a model prediction control technology is introduced, and an optimization control result is updated in real time, so that the influence of prediction errors and emergencies on the optimization control result is reduced, and voltage control of a full scheduling period is realized. According to the distributed new energy-based power distribution network active / reactive power coordinated optimization control method, active / reactive power coordinated optimization control of a power distribution network containing a distributed new energy / energy storage system is realized, the voltage out-of-limit problem caused by new energy and load fluctuation can be eliminated, voltage fluctuation can be reduced, and meanwhile, the operation economy of the power distribution network is considered.

Description

Technical Field

[0001] The present invention relates to the technical field of voltage control of distribution networks, and in particular to a method for coordinated optimization control of active and reactive power of distribution networks based on distributed renewable energy. Background Art

[0002] Due to the constraints of resources and environment and the constant changes in load types, in order to ensure the sustainability of resources and environment and improve energy utilization efficiency and reliability, smart grid has become the future development trend of power grid. According to the smart grid development plan of State Grid Corporation of China, the construction of smart grid in my country needs to greatly improve the ability of power grid to optimize resource allocation, which requires vigorously developing distributed new energy power generation resources, establishing a reasonable and efficient energy allocation method, and fully releasing the demand response capability of the user side to provide strong guarantee for the reliability of power supply for users. However, at present, in rural areas, mountainous areas, islands, highway service areas and other areas, due to being at the end of the distribution network with weak grid structure, and most of the loads in these areas have seasonal and periodic fluctuations, resulting in insufficient voltage support capacity, weak power supply capacity, poor power quality and other problems. In severe meteorological environments such as freezing rain and typhoons, the distribution network has many faults, low power supply reliability, and may even lose the main grid power supply. In addition, due to terrain restrictions, the repair work of the distribution network is extremely inconvenient, and it is difficult to restore power supply in a short time, which is easy to cause long-term power outages at the end of the distribution network. These problems are more prominent in the islands and mountainous areas along the southeast coast. In order to ensure the power supply of important loads such as hospitals, production and breeding, cable cars in tourist attractions and temples in these areas, diesel generators are currently used as backup power sources. However, diesel generators have high equipment and power generation costs, are noisy, and cause serious environmental pollution.

[0003] In the process of the development and construction of smart distribution networks, the increase in the access of intermittent distributed new energy sources such as photovoltaics and wind power, the deployment of energy storage and combined heat and power micro-turbines, the increase in active loads and the improvement of their observability and controllability have greatly enriched the operation and control means of distribution networks, which can effectively improve the power supply capacity and power quality of distribution networks. However, at the same time, it has also increased the complexity of distribution network operation and control. The control strategies and methods of traditional distribution networks are difficult to adapt to new changes, and active distribution networks with new energy, energy storage, and active load characteristics have emerged. Considering the output characteristics of new energy, the complexity of different types of energy storage components and controllable load control, as well as the continuous development and changes in power grid architecture and operation and control strategies, the optimization operation of distribution networks is becoming increasingly complex. With the increase in the access of new energy, the impact of its volatility on the voltage of distribution networks cannot be ignored. In distribution networks, active / reactive coordinated control has better voltage control effects than single active or reactive control, but the fluctuation of new energy (and load) will cause the voltage of distribution networks to exceed the limit. Summary of the invention

[0004] 1. Technical issues to be solved

[0005] In view of the problem that the fluctuation of the above-mentioned renewable energy (and load) may cause the voltage of the distribution network to exceed the limit, the present invention provides a method for coordinated optimization control of active and reactive power of the distribution network based on distributed renewable energy.

[0006] (II) Technical solution

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0008] A method for coordinated optimization control of active and reactive power in a distribution network based on distributed renewable energy comprises the following steps:

[0009] S1. Calculate the system power flow and node voltage-power sensitivity factor according to the current output status of the distributed power source, energy storage system and reactive power compensation equipment in the real-time monitoring system;

[0010] S2, considering the uncertainty of wind power, photovoltaic output and load, construct multiple scenarios through random sampling;

[0011] S3, establish the system active and reactive balance constraints in each time period, the active and reactive output constraints of various power sources, the coupling constraints between each time period, and the voltage constraints of each node in the first time period of each scheduling step;

[0012] S4. Based on the model predictive control technology, with the operation cost and the minimum node voltage deviation as the objective function, a real-time optimization control model is established, and a rolling solution is performed in multiple periods;

[0013] S5. The first time period dispatch result obtained in each optimization is used as the dispatch instruction for the system operation at the next moment to complete the real-time optimization control of the distribution network.

[0014] Furthermore, the wind power output prediction in step S1 adopts a classical model to fit the output power P of the entire wind farm at time t. WT (t) and the wind speed v at time t t relation:

[0015]

[0016] Where: v in Indicates the cut-in wind speed of the fan; v out Indicates the fan cut-out wind speed; v rate Represents the rated wind speed of the wind turbine; a0~a3 are fitting parameters based on the historical data of the wind farm.

[0017] Furthermore, the photovoltaic output prediction in step S1 combines actual meteorological data and photovoltaic output data to approximate the photovoltaic output short-term prediction error distribution to a normal distribution:

[0018]

[0019] Where: random variable X PV Represents the distributed photovoltaic prediction error. According to research, the mean value of the photovoltaic prediction error on the time axis, that is, the expected value, is zero; is the PV output prediction variance.

[0020] Furthermore, the load output prediction in step S1 uses a normal distribution function to describe the load prediction error, and its probability density function is shown as follows:

[0021]

[0022] Where: μ L is the mathematical expectation of load forecast; L 2 is the load forecast variance.

[0023] Furthermore, the step S2 takes the expected operating cost in each scenario as the objective function of the optimization problem:

[0024] minλV d,max +f cost

[0025] Where: f cost is the operating cost of the distribution network; λ is the voltage control coefficient; V d,max Represents the maximum deviation of each node voltage from the rated level;

[0026] The distribution network operating costs include the fuel costs of controllable power sources Start-Stop Cost Operation and maintenance costs of energy storage equipment ESS,ω (t) and the income from electricity purchase and sale, satisfying

[0027]

[0028] Where: ω represents different scenes; Ω is the set of scenes; π ω Represents the probability of each scene; P Buy (t), P Sale (t) are the buying and selling power of the distribution network respectively; U Buy (t), U Sale (t) represent the buying and selling prices of electricity in the distribution network respectively;

[0029] The relationship between the fuel cost of the controllable power source and the unit output can be described by the second-order Lagrangian function:

[0030]

[0031] Where: P iG,ω (t) is the controllable power output; a i 、b i 、c i are the fuel cost coefficients of the i-th controllable power source;

[0032]

[0033] Where: The startup cost of the controllable power supply; A 0-1 variable representing the power-on status of the controllable power supply;

[0034] The cost function of the energy storage unit includes investment depreciation cost and operation and maintenance cost:

[0035]

[0036] Where: R EES is the annual utilization rate; N EES The service life of the energy storage equipment; C is the investment and installation cost; M The operation and maintenance costs of energy storage equipment; are the charging and discharging power of the energy storage system respectively;

[0037]

[0038] Where: V N Indicates the rated voltage level; V i (t1) is the voltage amplitude at node i.

[0039] Furthermore, the constraints of step S3 are as follows:

[0040] (1) Power balance constraints

[0041]

[0042] Where P L,ω (t) is the active load in the distribution network; Q L,ω (t) is the reactive load in the distribution network; P PV,ω (t), Q PV,ω (t) is the active and reactive power of photovoltaic in the distribution network; P WT,ω (t), Q WT,ω (t) Active and reactive power of wind power in the distribution network; Q ESS,ω (t) is the reactive power of the energy storage device;

[0043] (2) Energy storage equipment operation constraints

[0044]

[0045] In the formula, SESS,ω (t) is the energy storage SOC value; are the charging and discharging efficiency of the energy storage device respectively; are the upper and lower limits of the energy storage charge state respectively; P ESS,ω (t) is the active power of the energy storage device; is the capacity of the inverter of the energy storage device;

[0046] (3) Controllable power supply operation constraints

[0047]

[0048] In the formula, They are the maximum and minimum outputs of the controllable power supply respectively; A 0-1 variable indicating the on / off status of the controllable power supply;

[0049] (4) Constraints on active and reactive output of photovoltaic and wind power

[0050] P PV,min (t)≤P PV (t)≤P PV,max (t)

[0051] Q PV,min (t)≤Q PV (t)≤Q PV,max (t)

[0052] P WT,min (t)≤P WT (t)≤P WT,max (t)

[0053] Q PV,min (t)≤Q PV (t)≤Q PV,max (t)

[0054] Where P PV,min (t) and P PV,max (t) are the lower and upper bounds of the active power of the photovoltaic power station node; Q PV,min (t) and Q PV,max (t) are the lower and upper bounds of the reactive power of the photovoltaic power station node; P WT,min (t) and P WT,max (t) are the lower and upper bounds of the active power of the wind power injection node; Q PV,min (t) and Q PV,max (t) are the lower and upper bounds of reactive power at the wind power injection node, respectively;

[0055] When photovoltaic cells are connected to the grid through an inverter, their reactive power upper limit is Q PV,max (t) can be obtained by the active power P of the photovoltaic power station node PV(t), photovoltaic grid voltage V PV (t) and the rated current of the photovoltaic power station inverter Joint decision:

[0056]

[0057] (5) Active and reactive power constraints at the grid connection point

[0058] P PCC (t) = P Buy (t)-P Sale (t)

[0059]

[0060] Where P PCC (t) is the active power at the connection point of the distribution network; is the upper limit of the grid-connected point power; Q PCC (t) is the reactive power absorbed by the distribution network from the large power grid; φ max Represents the maximum power factor angle allowed at the grid connection point;

[0061] (6) Node voltage constraints

[0062] V i,min ≤V i (t1)≤V i,max

[0063] ΔV i (t1)≤V i (t1)-V i (t0)

[0064] Among them, t1 is the starting time of the scheduling interval V i,max 、V i,min are the upper and lower limits of the node voltage respectively; V i (t0) is the voltage amplitude of node i in the current state; ΔV i (t1) represents the voltage amplitude change between time points t1 and t0, which can be obtained by solving the distribution network power flow equation:

[0065]

[0066] Where P Gi and Q Gi is the active power and reactive power of the power supply at node i; P Di and Q Di are the active and reactive power of the load at node i respectively; ΔP i and ΔQ i are the unbalanced amounts of active power and reactive power at node i, respectively; m is the number of nodes; V i and Vj are the voltage amplitudes of nodes i and j respectively; G ij , B ij are the real and imaginary parts of the admittance element between nodes i and j, respectively, θ ij is the voltage phase difference between nodes i and j.

[0067] Furthermore, the steps for implementing the power grid optimization control using model predictive control in step S4 are as follows:

[0068] 1) Let k = 0, and select the time interval of each control period [k, k+1] as ΔT = 15min;

[0069] 2) According to the time scale of ultra-short-term forecasting of renewable energy and load, the intraday forecast period and the optimization control period are both 4 hours, that is, N P =N C =16; N C For a finite time domain, N P For the prediction time domain;

[0070] 3) Ultra-short-term forecasting of new energy power and load;

[0071] 4) Establish an intraday optimization control model for the distribution network;

[0072] 5) In the entire control time domain N C On the other hand, the optimization control model of the power grid is used to solve the optimization problem of the distribution network and obtain the entire control time domain N C Control instructions for accessing the power grid;

[0073] 6) Sending the control instructions for the current k period to the dispatchable components in the power grid;

[0074] 7) Set [k, k+1], measure the current system state information, update the prediction information and the initial state information in the power grid intraday optimization control model, and go to step 4).

[0075] (III) Beneficial effects

[0076] The beneficial effects of the present invention are as follows: a method for coordinated optimization control of active and reactive power of a distribution network based on distributed renewable energy, which introduces a multi-scenario stochastic modeling method to reduce the impact of load and renewable energy output volatility and uncertainty factors, and ensure the robustness of the optimization control strategy; the model predictive control technology is introduced to update the optimization control results in real time to reduce the impact of prediction errors and emergencies on the optimization control results, and realize voltage control of the entire scheduling cycle; the present invention realizes the coordinated optimization control of active and reactive power of a distribution network containing distributed renewable energy / energy storage systems, which can eliminate the voltage over-limit problem caused by renewable energy and load fluctuations and reduce voltage fluctuations, while taking into account the economy of the distribution network operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0078] Figure 1 This is a flow chart of the control method of the present invention. DETAILED DESCRIPTION

[0079] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0080] Combination Figure 1 , a method for coordinated optimization control of active and reactive power in a distribution network based on distributed renewable energy, comprising the following steps:

[0081] S1. Calculate the system power flow and node voltage-power sensitivity factor based on the current output status of various components (including wind power, photovoltaic power, and load) in the real-time monitoring system;

[0082] (1) Wind power output forecast

[0083] The output level of the wind turbine is mainly determined by the wind speed. The greater the wind speed, the greater the kinetic energy of the air per unit volume, and the greater the power. The present invention uses a classic model to fit the output power P of the entire wind farm at time t WT (t) and the wind speed v at time t t relation:

[0084]

[0085] Where: v in Indicates the cut-in wind speed of the fan; v out Indicates the fan cut-out wind speed; v rate Indicates the rated wind speed of the wind turbine; a0~a3 are fitting parameters, which can be obtained by fitting based on the historical data of the wind farm.

[0086] (2) Photovoltaic output forecast

[0087] The analysis of the short-term forecast error of photovoltaic output is carried out through random sampling in the error sample data. Combined with the actual meteorological data and photovoltaic output data, the distribution of the short-term forecast error of photovoltaic output is approximated to a normal distribution:

[0088]

[0089] Where: random variable X PV Represents the distributed photovoltaic prediction error. According to research, the mean value of the photovoltaic prediction error on the time axis, that is, the expected value, is zero; is the PV output prediction variance.

[0090] (3) Load output prediction

[0091] Due to the limitations of load forecasting technology itself and the problems of data quality and data comprehensiveness, there are always errors in existing load forecasting. In the process of uncertainty modeling, in order to describe the uncertainty of load, the normal distribution function is often used to describe the load forecast error. Its probability density function is shown as follows:

[0092]

[0093] Where: μ L is the mathematical expectation of load forecast; σ L 2 is the load forecast variance.

[0094] S2, considering the uncertainty of wind power, photovoltaic output and load, construct multiple scenarios through random sampling;

[0095] Since wind power, photovoltaic output and load are described using a multi-scenario method, the expected operating cost under each scenario is taken as the objective function of the optimization problem:

[0096] minλV d,max +f cost

[0097] Where: f cost is the operating cost of the distribution network; λ is the voltage control coefficient; V d,max Represents the maximum deviation of each node voltage from the rated level. Among them, λ is an optimization control model parameter. The larger its value, the smaller the deviation between the node voltage and the rated value in the dispatch result will be. In the actual operation of the distribution network, different λ values ​​can be selected according to the specific requirements of the voltage level.

[0098] The distribution network operating costs include the fuel costs of controllable power sources Start-Stop Cost Operation and maintenance costs of energy storage equipment ESS,ω (t) and the income from electricity purchase and sale, satisfying

[0099]

[0100] Where: ω represents different scenes; Ω is the set of scenes; π ω Represents the probability of each scene; P Buy (t), P Sale (t) are the buying and selling power of the distribution network respectively; U Buy (t), U Sale (t) represent the buying and selling prices of electricity in the distribution network respectively.

[0101] The relationship between the fuel cost of the controllable power source and the unit output can be described by the second-order Lagrangian function:

[0102]

[0103] Where: P i G,ω (t) is the controllable power output; a i 、b i 、c i are the fuel cost coefficients of the i-th controllable power source.

[0104]

[0105] Where: The startup cost of the controllable power supply; A 0-1 variable representing the power-on status of the controllable power supply.

[0106] The cost function of the energy storage unit includes investment depreciation cost and operation and maintenance cost:

[0107]

[0108] Where: R EES is the annual utilization rate; N EES The service life of the energy storage equipment; C is the investment and installation cost; M The operation and maintenance costs of energy storage equipment; are the charging and discharging power of the energy storage system respectively.

[0109]

[0110] Where: V N Indicates the rated voltage level; V i (t1) is the voltage amplitude at node i.

[0111] S3, establish the system active and reactive balance constraints in each time period, the active and reactive output constraints of various power sources, the coupling constraints between each time period, and the voltage constraints of each node in the first time period of each scheduling step;

[0112] (1) Power balance constraints:

[0113]

[0114] Where P L,ω (t) is the active load in the distribution network; Q L,ω (t) is the reactive load in the distribution network; P PV,ω (t), Q PV,ω (t) is the active and reactive power of photovoltaic in the distribution network; P WT,ω (t), Q WT,ω (t) Active and reactive power of wind power in the distribution network; Q ESS,ω (t) is the reactive power of the energy storage device.

[0115] (2) Energy storage equipment operation constraints

[0116]

[0117]

[0118] In the formula, S ESS,ω (t) is the energy storage SOC value; are the charging and discharging efficiency of the energy storage device respectively; are the upper and lower limits of the energy storage charge state respectively; P ESS,ω (t) is the active power of the energy storage device; is the capacity of the inverter of the energy storage device.

[0119] (3) Controllable power supply operation constraints

[0120]

[0121] In the formula, They are the maximum and minimum outputs of the controllable power supply respectively; A 0-1 variable indicating the on / off status of the controllable power supply.

[0122] (4) Constraints on active and reactive output of photovoltaic and wind power

[0123] P PV,min (t)≤P PV (t)≤P PV,max (t)

[0124] Q PV,min (t)≤Q PV (t)≤Q PV,max (t)

[0125] P WT,min (t)≤P WT (t)≤PWT,max (t)

[0126] Q PV,min (t)≤Q PV (t)≤Q PV,max (t)

[0127] Where P PV,min (t) and P PV,max (t) are the lower and upper bounds of the active power of the photovoltaic power station node; Q PV,min (t) and Q PV,max (t) are the lower and upper bounds of the reactive power of the photovoltaic power station node; P WT,min (t) and P WT,max (t) are the lower and upper bounds of the active power of the wind power injection node; Q PV,min (t) and Q PV,max (t) are the lower and upper bounds of reactive power at the wind power injection node, respectively.

[0128] When photovoltaic cells are connected to the grid through an inverter, their reactive power upper limit is Q PV,max (t) can be obtained by the active power P of the photovoltaic power station node PV (t), photovoltaic grid voltage V PV (t) and the rated current of the photovoltaic power station inverter Joint decision:

[0129]

[0130] (5) Active and reactive power constraints at the grid connection point

[0131] P PCC (t) = P Buy (t)-P Sale (t)

[0132]

[0133] Where P PCC (t) is the active power at the connection point of the distribution network; is the upper limit of the grid-connected point power; Q PCC (t) is the reactive power absorbed by the distribution network from the large power grid; φ max Represents the maximum power factor angle allowed at the grid connection point.

[0134] (6) Node voltage constraints

[0135] V i,min ≤V i (t1)≤V i,max

[0136] ΔV i (t1)≤V i(t1)-V i (t0)

[0137] This formula indicates that only the node voltage constraint in the first period is considered in each step of the scheduling model. Where t1 is the starting time V of the scheduling interval. i,max 、V i,min are the upper and lower limits of the node voltage respectively; V i (t0) is the voltage amplitude of node i in the current state; ΔV i (t1) represents the voltage amplitude change between time points t1 and t0, which can be obtained by solving the distribution network power flow equation:

[0138]

[0139] Where P Gi and Q Gi is the active power and reactive power of the power supply at node i; P Di and Q Di are the active and reactive power of the load at node i respectively; ΔP i and ΔQ i are the unbalanced amounts of active power and reactive power at node i, respectively; m is the number of nodes; V i and V j are the voltage amplitudes of nodes i and j respectively; G ij , B ij are the real and imaginary parts of the admittance element between nodes i and j, respectively, θ ij is the voltage phase difference between nodes i and j.

[0140] Performing Taylor expansion and ignoring higher-order terms, we can obtain the following matrix expression:

[0141]

[0142] Where: Δθ is the node voltage phase angle change matrix; ΔV is the node voltage amplitude change matrix; ΔP is the node active power change matrix; ΔQ is the node reactive power change matrix; J is the Jacobian matrix of the conventional power flow model.

[0143] Let the inverse matrix of J be J -1 :

[0144]

[0145] Then the sensitivity matrix of voltage amplitude and injected active and reactive power in the distribution network can be obtained as

[0146] ΔV=S 21 ΔP+S 22 ΔQ

[0147] Where S21 It is called voltage-active power sensitivity factor; S 22 It is called the voltage-reactive power sensitivity factor.

[0148] S4. Based on the Model Predictive Control (MPC) technology, with the operation cost and the minimum node voltage deviation as the objective function, a real-time optimization control model is established, and a rolling solution is performed over multiple periods;

[0149] The implementation steps of power grid optimization control using model predictive control are as follows:

[0150] 1) Let k = 0, and select the time interval of each control period [k, k+1] as ΔT = 15min;

[0151] 2) According to the time scale of ultra-short-term forecasting of renewable energy and load, the intraday forecast period and the optimization control period are both 4 hours, that is, N P =N C =16; N C For a finite time domain, N P For the prediction time domain;

[0152] 3) Ultra-short-term forecasting of new energy power and load;

[0153] 4) Establish an intraday optimization control model for the distribution network;

[0154] 5) In the entire control time domain N C On the other hand, the optimization control model of the power grid is used to solve the optimization problem of the distribution network and obtain the entire control time domain N C Control instructions for accessing the power grid;

[0155] 6) Sending the control instructions for the current k period to the dispatchable components in the power grid;

[0156] 7) Set [k, k+1], measure the current system state information, update the prediction information and the initial state information in the power grid intraday optimization control model, and go to step 4).

[0157] S5. The first time period dispatch result obtained in each optimization is used as the dispatch instruction for the system operation at the next moment to complete the real-time optimization control of the distribution network.

[0158] Compared with the single-scenario deterministic method and the active / reactive independent optimization control method, the voltage control effect and economy of the distribution network operation can be optimized under the multi-scenario active / reactive coordinated optimization control strategy.

[0159] In summary, the embodiment of the present invention is based on the active and reactive coordinated optimization control method of the distribution network of distributed renewable energy, introduces a multi-scenario stochastic modeling method, reduces the impact of load and renewable energy output volatility and uncertainty factors, and ensures the robustness of the optimization control strategy; introduces model predictive control technology to update the optimization control results in real time to reduce the impact of prediction errors and emergencies on the optimization control results, and realizes voltage control of the entire scheduling cycle; the present invention realizes the active / reactive coordinated optimization control of the distribution network containing distributed renewable energy / energy storage systems, which can eliminate the voltage over-limit problem caused by renewable energy and load fluctuations and reduce voltage fluctuations, while taking into account the economy of the distribution network operation.

[0160] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for coordinated optimization control of active and reactive power in a distribution network based on distributed renewable energy, characterized in that: The following steps are involved: S1. Calculate the system power flow and node voltage-power sensitivity factor according to the current output status of the distributed power source, energy storage system and reactive power compensation equipment in the real-time monitoring system; S2, considering the uncertainty of wind power, photovoltaic output and load, construct multiple scenarios through random sampling; S3, establish the system active and reactive balance constraints in each time period, the active and reactive output constraints of various power sources, the coupling constraints between each time period, and the voltage constraints of each node in the first time period of each scheduling step; S4. Based on the model predictive control technology, with the operation cost and the minimum node voltage deviation as the objective function, a real-time optimization control model is established, and a rolling solution is performed in multiple periods; S5. The first time period dispatch result obtained in each optimization is used as the dispatch instruction for the system operation at the next moment to complete the real-time optimization control of the distribution network.

2. The method for coordinated optimization control of active and reactive power in a distribution network based on distributed renewable energy as claimed in claim 1, characterized in that: The wind power output prediction in step S1 adopts a classical model to fit the output power P of the entire wind farm at time t. WT (t) and the wind speed v at time t t relation: Where: v in Indicates the cut-in wind speed of the fan; v out Indicates the fan cut-out wind speed; v rate Represents the rated wind speed of the wind turbine; a0~a3 are fitting parameters based on the historical data of the wind farm.

3. The method for coordinated optimization control of active and reactive power in a distribution network based on distributed renewable energy as claimed in claim 2, characterized in that: The photovoltaic output prediction in step S1 combines the actual meteorological data and photovoltaic output data to approximate the photovoltaic output short-term prediction error distribution to a normal distribution: Where: random variable X PV Represents the distributed photovoltaic prediction error. According to research, the mean value of the photovoltaic prediction error on the time axis, that is, the expected value, is zero; is the PV output prediction variance.

4. The method for coordinated optimization control of active and reactive power in a distribution network based on distributed renewable energy as claimed in claim 3, characterized in that: The load output prediction in step S1 uses a normal distribution function to describe the load prediction error, and its probability density function is shown as follows: Where: μ L is the mathematical expectation of load forecast; σ L 2 is the load forecast variance.

5. The method for coordinated optimization control of active and reactive power in a distribution network based on distributed renewable energy as claimed in claim 1, characterized in that: The step S2 takes the expected operating cost in each scenario as the objective function of the optimization problem: minλV d,max +f cost (4) Where: f cost is the operating cost of the distribution network; λ is the voltage control coefficient; V d,max Represents the maximum deviation of each node voltage from the rated level; The distribution network operating costs include the fuel costs of controllable power sources Start-Stop Cost Operation and maintenance costs of energy storage equipment ESS,ω (t) and the income from electricity purchase and sale, satisfying Where: ω represents different scenes; Ω is the set of scenes; π ω Represents the probability of each scene; P Buy (t), P Sale (t) are the buying and selling power of the distribution network respectively; U Buy (t), U Sale (t) represent the buying and selling prices of electricity in the distribution network respectively; The relationship between the fuel cost of the controllable power source and the unit output can be described by the second-order Lagrangian function: Where: P i G,ω (t) is the controllable power output; a i 、b i 、c i are the fuel cost coefficients of the i-th controllable power source; Where: The startup cost of the controllable power supply; A 0-1 variable representing the power-on status of the controllable power supply; The cost function of the energy storage unit includes investment depreciation cost and operation and maintenance cost: Where: R EES is the annual utilization rate; N EES The service life of the energy storage equipment; C is the investment and installation cost; M The operation and maintenance costs of energy storage equipment; are the charging and discharging power of the energy storage system respectively; Where: V N Indicates the rated voltage level; V i (t1) is the voltage amplitude at node i.

6. The method for coordinated optimization control of active and reactive power in a distribution network based on distributed renewable energy as claimed in claim 1, characterized in that: The constraints of step S3 are as follows: (1) Power balance constraints Where P L,ω (t) is the active load in the distribution network; Q L,ω (t) is the reactive load in the distribution network; P PV,ω (t), Q PV,ω (t) is the active and reactive power of photovoltaic in the distribution network; P WT,ω (t), Q WT,ω (t) Active and reactive power of wind power in the distribution network; Q ESS,ω (t) is the reactive power of the energy storage device; (2) Energy storage equipment operation constraints In the formula, S ESS,ω (t) is the energy storage SOC value; are the charging and discharging efficiency of the energy storage device respectively; are the upper and lower limits of the energy storage charge state respectively; P ESS,ω (t) is the active power of the energy storage device; is the capacity of the inverter of the energy storage device; (3) Controllable power supply operation constraints In the formula, They are the maximum and minimum outputs of the controllable power supply respectively; A 0-1 variable indicating the on / off status of the controllable power supply; (4) Constraints on active and reactive output of photovoltaic and wind power P PV,min (t)≤P PV (t)≤P PV,max (t) (20) Q PV,min (t)≤Q PV (t)≤Q PV,max (t) (21) P WT,min (t)≤P WT (t)≤P WT,max (t) (22) Q PV,min (t)≤Q PV (t)≤Q PV,max (t) (23) Where P PV,min (t) and P PV,max (t) are the lower and upper bounds of the active power of the photovoltaic power station node; Q PV,min (t) and Q PV ,max (t) are the lower and upper bounds of the reactive power of the photovoltaic power station node; P WT,min (t) and P WT,max (t) are the lower and upper bounds of the active power of the wind power injection node; Q PV,min (t) and Q PV,max (t) are the lower and upper bounds of reactive power at the wind power injection node, respectively; When photovoltaic cells are connected to the grid through an inverter, their reactive power upper limit is Q PV,max (t) can be obtained by the active power P of the photovoltaic power station node PV (t), photovoltaic grid voltage V PV (t) and the rated current of the photovoltaic power station inverter Joint decision: (5) Active and reactive power constraints at the grid connection point P PCC (t)=P Buy (t)-P Sale (t) (25) Where P PCC (t) is the active power at the connection point of the distribution network; is the upper limit of the grid-connected point power; Q PCC (t) is the reactive power absorbed by the distribution network from the large power grid; φ max Represents the maximum power factor angle allowed at the grid connection point; (6) Node voltage constraints 5 i,min ≤V i (t1)≤V i,max (29) ΔV i (t1)≤V i (t1)-V i (t0) (30) Among them, t1 is the starting time of the scheduling interval V i,max 、V i,min are the upper and lower limits of the node voltage respectively; V i (t0) is the voltage amplitude of node i in the current state; ΔV i (t1) represents the voltage amplitude change between time points t1 and t0, which can be obtained by solving the distribution network power flow equation: Where P Gi and Q Gi is the active power and reactive power of the power supply at node i; P Di and Q Di are the active and reactive power of the load at node i respectively; ΔP i and ΔQ i are the unbalanced amounts of active power and reactive power at node i, respectively; m is the number of nodes; V i and V j are the voltage amplitudes of nodes i and j respectively; G ij , B ij are the real and imaginary parts of the admittance element between nodes i and j, respectively, θ ij is the voltage phase difference between nodes i and j.

7. A method for coordinated optimization control of active and reactive power in a distribution network based on distributed renewable energy as claimed in claim 6, characterized in that: The implementation steps of the power grid optimization control using model predictive control in step S4 are as follows: 1) Let k = 0, and select the time interval of each control period [k, k+1] as ΔT = 15min; 2) According to the time scale of ultra-short-term forecasting of renewable energy and load, the intraday forecast period and the optimization control period are both 4 hours, that is, N P =N C =16; N C is a finite time domain, N P For the prediction time domain; 3) Ultra-short-term forecasting of new energy power and load; 4) Establish an intraday optimization control model for the distribution network; 5) In the entire control time domain N C On the other hand, the optimization control model of the power grid is used to solve the optimization problem of the distribution network and obtain the entire control time domain N C Control instructions for accessing the power grid; 6) Sending the control instructions for the current k period to the dispatchable components in the power grid; 7) Set [k, k+1], measure the current system state information, update the prediction information and the initial state information in the power grid intraday optimization control model, and go to step 4).

Images