Intelligent farm virtual power plant model containing agricultural irrigation load and optimal scheduling method

Abstract

The invention discloses an agricultural irrigation load-containing intelligent farm virtual power plant model and an optimal scheduling method, and belongs to the field of power systems, and the model comprises the steps of constructing a typical scene based on an affinity propagation clustering algorithm, constructing an agricultural-power coupling relation model, and constructing a farm virtual power plant operation cost calculation model. The optimal scheduling method can achieve the operation effects of single-field stabilization, multi-field peak shifting, production guaranteeing and overall optimization, calculates the optimal regulation and scheduling scheme of the flexible resources of the intelligent farm virtual power plant according to the obtained typical scene operation random optimization strategy by using the regulation and control total cost model of the flexible resources of the intelligent farm virtual power plant, and achieves the optimal regulation and scheduling of the flexible resources of the intelligent farm virtual power plant. And obtaining a day-ahead scheduling strategy. According to the method, the regulation and control level of the intelligent farm virtual power plant can be effectively improved, distributed energy resources are effectively integrated and optimized, the energy utilization efficiency is improved, and the farm virtual power plant is promoted to develop in a more environment-friendly, efficient and flexible direction, so that the sustainable development target of an electric power-agriculture coupling system is achieved.

Description

Technical Field

[0001] The present invention belongs to the field of electric power systems, and in particular relates to an intelligent farm virtual power plant model including agricultural irrigation loads and an optimization scheduling method. Background Art

[0002] Virtual power plant is an innovative energy management system. It realizes the aggregation and coordinated optimization of distributed energy resources such as distributed power sources (such as renewable energy, energy storage system), controllable load, electric vehicles, etc. through advanced information and communication technology and software system. The core of virtual power plant lies in "aggregation" and "communication". It can connect various distributed energy resources and realize the aggregation and coordinated optimization of these resources through software system to form a special "power plant". Virtual power plant has the functions of traditional power plant, can achieve accurate automatic response, and its unit characteristic curve can also simulate conventional generator set. Virtual power plant can monitor and predict various factors such as energy demand and market price in real time, and flexibly dispatch and coordinate distributed energy resources according to these factors. For example, it can quickly increase output when power demand increases, and reduce output or store excess power when power supply is in excess. As a special power plant, virtual power plant participates in the power market and power grid operation. It can serve as a "positive power plant" to supply power to the system for peak load regulation, and can also serve as a "negative power plant" to increase load consumption and coordinate system valley filling.

[0003] Smart farm virtual power plants usually have a large number of distributed energy resources, such as solar photovoltaic panels, wind turbines, smart irrigation loads, etc.

[0004] The existing dispatching methods for farm virtual power plants still have the following three major difficulties that need to be overcome: 1) The power dispatching plan ignores the agricultural production plan, and there is a risk of crop yield reduction. There is a lack of overall consideration of power supply and agricultural production supply, the model of farm output and electricity consumption is unclear, and there is a lack of specific modeling of the relationship chain of irrigation electricity-irrigation water-crop yield; 2) The dynamic change process of farmland soil moisture is not clearly modeled. The irrigation load dispatching model of existing research is too simple and lacks a scientific description of the specific demand for irrigation water. For example, it only considers electric water extraction and does not consider the impact of transpiration, making it difficult to effectively guide farmland irrigation and equipment electricity consumption; 3) The production and operation of farm virtual power plants are affected by multiple uncertainties such as precipitation. How to ensure the safe operation and production supply of farm virtual power plants under multiple uncertainties requires further in-depth research. Summary of the invention

[0005] In response to the above problems, the present invention provides a modeling and optimization scheduling method for a smart farm virtual power plant with agricultural irrigation load, which can effectively improve the regulation level of the smart farm virtual power plant, effectively integrate and optimize distributed energy resources, improve energy utilization efficiency, reduce energy waste, and at the same time alleviate the risks brought by the multiple uncertainties of the agricultural production process to agricultural production supply and virtual power plant operation.

[0006] The technical solution adopted by the present invention is as follows: A method for constructing a virtual power plant model of an intelligent farm including agricultural irrigation loads comprises the following steps:

[0007] S1: Construct a typical scenario based on the nearest neighbor propagation clustering algorithm: collect weather and electricity price forecast information of the smart farm virtual power plant, and use the clustering algorithm to obtain the typical scenario for the day-ahead dispatch of the virtual power plant, including the output information of wind power equipment, the output information of photovoltaic equipment, the fluctuation information of market electricity prices and the fluctuation information of precipitation;

[0008] S2: Constructing the agriculture-electricity coupling relationship model: including constructing the crop water demand model, soil moisture change model, crop yield model and irrigation water-electricity conversion model.

[0009] The crop water requirement model includes crop coefficient, crop real water requirement parameter, reference crop water requirement parameter, saturated water vapor pressure curve slope parameter, surface net radiation parameter, soil heat flux parameter, psychrometric constant, daily average temperature parameter, wind speed parameter at 2 meters above the surface, saturated water vapor pressure parameter, and actual water vapor pressure parameter;

[0010] The soil moisture variation model includes soil moisture content parameters, effective rainfall parameters, shallow groundwater daily recharge parameters, irrigation parameters, and soil moisture evaporation parameters.

[0011] The crop yield model includes crop evapotranspiration parameters, soil moisture correction coefficient, average soil moisture content parameters in the crop root activity layer, soil wilting coefficient, and crop water production function.

[0012] The irrigation water-to-electricity conversion model includes a parameter of farmland irrigation electricity consumption multiplied by a water-to-electricity conversion coefficient;

[0013] S3: constructing a farm virtual power plant operation cost calculation model, which is used to calculate the total cost of the farm virtual power plant's electricity consumption throughout the day and the potential production reduction. The farm virtual power plant operation cost calculation model includes constructing a virtual power plant flexibility resource model and a virtual power plant agriculture-power operation cost model;

[0014] The virtual power plant flexibility resource model includes a wind and solar power model, an energy storage device model, a reactive power compensation SVG device model, and an irrigation load model. It models the operation principle of all flexibility resources in the virtual power plant, builds the physical basis of the virtual power plant control strategy, and uses power balance constraints to ensure the safety of virtual power plant operation.

[0015] The virtual power plant agriculture-electricity operation cost model calculates the virtual power plant electricity cost, agricultural production reduction cost, and virtual power plant electricity sales revenue, and then comprehensively calculates the total benefit of the virtual power plant operation.

[0016] Furthermore, in step S1, a clustering algorithm is used to obtain a typical scenario for the day-ahead dispatch of a virtual power plant, as shown in formulas (1)-(4):

[0017] Formula (1) is the method for calculating the similarity matrix elements;

[0018]

[0019] Where: s(m,n) is the similarity matrix element, the non-diagonal elements are Euclidean distances; m, m', n, n' are the point numbers; x m 、x n are the uncertain parameters respectively;

[0020] Formula (2) is the method for calculating the elements of the attraction matrix;

[0021]

[0022] In the formula: r(m,n) and a(m,n') are the elements of the attraction matrix and the belonging matrix respectively;

[0023] Formula (3) is the calculation method of the attribution matrix elements;

[0024]

[0025] Formula (4) attenuation iterative update calculation;

[0026]

[0027] Where λ represents the damping coefficient and s represents the number of iterations.

[0028] Furthermore, in step S2, formulas (5)-(6) are crop water requirement models,

[0029] ET=K C ×ET R (5)

[0030] Where: K C is the crop coefficient, ET is the actual water requirement of the crop, and ETR is the reference crop water requirement;

[0031]

[0032] Where: Δ is the slope of the saturated water vapor pressure curve; R n is the net radiation of the surface; G is the soil heat flux; γ is the psychrometric constant; T mean is the daily average temperature; u2 is the wind speed at 2 meters above the ground; e s is the saturated water vapor pressure; e a is the actual water vapor pressure;

[0033] Formula (7) is the soil moisture change model,

[0034] W t =W t-1 +P 0,t +K t +M t -ET t (7)

[0035] Where: t is the time; W t is the soil water content at time t; W t-1 is the soil moisture content at the previous moment; P 0,t is the effective rainfall in this period; K t is the shallow groundwater recharge during this period; M t is the irrigation amount during this period; ET t is the soil water evaporation during this period;

[0036] Formulas (8)-(11) are crop yield models.

[0037] ET A =K s ×ET(8)

[0038] Where: ET A K is the crop evapotranspiration; s is the soil moisture correction factor;

[0039]

[0040] Where: θ is the average soil moisture content in the crop root activity layer during the calculation period; θ f is the upper limit of the optimal soil moisture; θ wp is the soil wilting coefficient;

[0041]

[0042] Where: Y a , Y max , E.T. A,k , E.T.k They represent the actual crop yield, the maximum yield, the sum of the actual transpiration in production cycle k, and the sum of the actual daily crop water demand in production cycle k; B k is the moisture sensitivity empirical coefficient;

[0043]

[0044] Where: D k,j is the potential yield reduction rate of each farmland per day; j is the day sequence, Y a,k,j , Y max,k,j , E.T. A,k,j They represent the equivalent daily output value of farmland, the maximum potential daily output value, and the evaporation and transpiration of crops per day respectively;

[0045] Formula (12) is the irrigation water-to-electricity conversion model,

[0046]

[0047] Where: is the irrigation water volume of farmland i during this period; is the electricity consumption for irrigation of farmland i during this period, T C is the water-to-electricity conversion coefficient.

[0048] Furthermore, in step S3, formula (13) is the wind and solar power model,

[0049]

[0050] Where: are the maximum output power of wind and solar power sources at time t, and the upper limits of wind and solar power sources are the predicted output values ​​after uncertainty processing; P t WT , P t PV are wind power / photovoltaic output values ​​respectively;

[0051] Formulas (14)-(16) are energy storage device models.

[0052]

[0053] Where: is the remaining energy of the battery; Divided into the initial, minimum and maximum remaining energy of the battery; It is the charge / discharge mark of the battery;

[0054]

[0055] Where: P t ESS,in , P tESS,out are the battery charging / discharging power respectively; They are the upper limits of battery charging / discharging power respectively;

[0056]

[0057] Where: η in , η out , γ ESS are the charging / discharging efficiency / dissipation rate of the battery respectively; T is the total number of moments; ΔT is the total time period;

[0058] Formula (17) is the reactive power compensation SVG equipment model,

[0059]

[0060] Where: is the maximum reactive compensation power of the SVG device; Contribute to SVG devices;

[0061] Formula (18) is the irrigation load model,

[0062]

[0063] Where: is the maximum power of the irrigation load; P t well is the operating power of the irrigation load;

[0064] Formula (19) is the power balance constraint,

[0065]

[0066] Where: Purchase electricity for virtual power plants; Purchase electricity for the virtual power plant; P fix is a fixed active load; is the reactive power balance of the power grid; Q fix is a fixed reactive load;

[0067] Formula (20) is the soil moisture constraint for agricultural production;

[0068] W i min ≤W i t ≤W i max (20)

[0069] Where: W i t is the soil moisture of farmland i at time t; W imin , W i max are the lower and upper limits of soil moisture respectively; the agriculture-electricity operation cost model of the virtual power plant is expressed by formula (21):

[0070]

[0071] Where: C is the total cost; The price of electricity purchased from the grid; The price of electricity sold to the grid; Purchase electricity for virtual power plants; Purchase electricity for virtual power plants; D i Y is the potential daily yield reduction rate in this growth stage; m i is the estimated total output value of a single farmland; N is the total number of farmlands.

[0072] Another object of the present invention is to provide a smart farm virtual power plant model system including agricultural irrigation loads, comprising: a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the computer program, when executed by the processor, implements a method for constructing a smart farm virtual power plant model including agricultural irrigation loads as described above.

[0073] Another object of the present invention is to provide an optimization scheduling method for a smart farm virtual power plant model system including agricultural irrigation loads, so as to achieve the operation effects of single-field leveling, multi-field peak shifting, production guarantee, and overall optimization. The method is as follows: first, according to the typical scenario generation technology based on the neighbor propagation clustering algorithm, the typical scenario for the day-ahead scheduling reference of the virtual power plant is obtained, and the water demand and soil moisture requirements of the crops on the day are calculated according to the parameters in the typical prediction scenario; then, the agriculture-electricity coupling relationship model and the farm virtual power plant operation cost calculation model are constructed, the topological structure of the virtual power plant operation is built, and the power balance constraints and agricultural production soil moisture constraints of the virtual power plant operation are constructed; finally, the total cost model for the regulation of the flexible resources of the smart farm virtual power plant is used, and the random optimization strategy is run according to the typical scenario obtained, and the optimal regulation and scheduling plan for the flexible resources of the smart farm virtual power plant is calculated to obtain the day-ahead scheduling strategy;

[0074] Among them, formula (22) shows the total cost model of regulating the flexibility resources of the smart farm virtual power plant, including the electricity cost of the virtual power plant in a typical scenario, the cost of agricultural production reduction in a typical scenario, and the electricity sales income of the virtual power plant in a typical scenario.

[0075]

[0076] Where: U is the number of typical scenes after clustering; D u i is the potential daily reduction rate of the typical scenario in this growth stage, Ym,u i is the estimated total output value of a single farmland under a typical scenario; p i (u) is the probability of occurrence of a typical scenario; They are the electricity purchased and sold at time t, are the electricity purchase price and electricity sales price at time t respectively.

[0077] Another object of the present invention is to provide an optimization scheduling system for an intelligent farm virtual power plant including agricultural irrigation loads, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program implements the optimization scheduling method described above when executed by the processor.

[0078] The advantages and beneficial effects of the present invention are as follows: 1) It can intelligently dispatch energy supply according to the actual needs of the farm to avoid energy surplus or shortage, thereby saving energy costs; 2) It helps to promote the widespread application of renewable energy, reduce fossil energy consumption and pollutant emissions. In a farm environment, it helps to achieve green development and sustainable management of agriculture; 3) It can ensure the stable operation of the agricultural production process according to the farm's production plan and energy needs. By integrating and optimizing energy resources, the virtual power plant can also provide a stable power supply for the farm to ensure the continuity and efficiency of agricultural production; 4) It can realize real-time perception, response and adjustment of farm energy supply and demand, and improve the flexibility and adaptability of agricultural production; 5) The smart farm virtual power plant can participate in the power market and power grid operation as a special power plant. In a farm environment, this means that the farm can participate in power market transactions through the virtual power plant platform with its own distributed energy resources, thereby obtaining additional benefits. By participating in electricity market transactions, farms can also better understand market dynamics and price changes, providing strong support for future energy investment decisions; 6) Provide physical support for the modeling of future optimization planning and operation methods of other virtual power plants containing agricultural irrigation equipment, and promote the development of farm virtual power plants in a more environmentally friendly, efficient and flexible direction, thereby achieving the sustainable development goals of the power-agriculture coupling system. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] Figure 1 Flowchart of clustering algorithm;

[0080] Figure 2 The overall flow chart for solving the optimal operation of farm virtual power plants;

[0081] Figure 3 Irrigation plan diagram before optimization;

[0082] Figure 4 Optimized irrigation plan diagram. DETAILED DESCRIPTION

[0083] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0084] Example 1

[0085] A method for constructing a smart farm virtual power plant model including agricultural irrigation loads comprises the following steps:

[0086] S1: Construct a typical scenario based on the nearest neighbor propagation clustering algorithm: collect weather and electricity price forecast information of the smart farm virtual power plant, and use the clustering algorithm to obtain the typical scenario of the virtual power plant's day-ahead dispatch reference. The process of the clustering algorithm is as follows: Figure 1 As shown, it includes the output information of wind power equipment, the output information of photovoltaic equipment, the fluctuation information of market electricity price and the fluctuation information of precipitation;

[0087] Among them, the clustering algorithm is used to obtain the typical scenario of virtual power plant day-ahead dispatch reference, as shown in formulas (1)-(4):

[0088] Formula (1) is the method for calculating the similarity matrix elements;

[0089]

[0090] Where: s(m,n) is the similarity matrix element, the non-diagonal elements are Euclidean distances; m, m', n, n' are the point numbers; x m 、x n are the uncertain parameters respectively;

[0091] Formula (2) is the method for calculating the elements of the attraction matrix;

[0092]

[0093] In the formula: r(m,n) and a(m,n') are the elements of the attraction matrix and the belonging matrix respectively;

[0094] Formula (3) is the calculation method of the attribution matrix elements;

[0095]

[0096] Formula (4) attenuation iterative update calculation;

[0097]

[0098] Where λ represents the damping coefficient and s represents the number of iterations.

[0099] S2: Constructing the agriculture-electricity coupling relationship model: including constructing the crop water demand model, soil moisture change model, crop yield model and irrigation water-electricity conversion model.

[0100] Formulas (5) and (6) are crop water requirement models.

[0101] ET=K C ×ET R (5)

[0102] Where: K C is the crop coefficient, ET is the actual water requirement of the crop, and ET R is the reference crop water requirement;

[0103]

[0104] Where: Δ is the slope of the saturated water vapor pressure curve; R n is the net radiation of the surface; G is the soil heat flux; γ is the psychrometric constant; T mean is the daily average temperature; u2 is the wind speed at 2 meters above the ground; e s is the saturated water vapor pressure; e a is the actual water vapor pressure;

[0105] Formula (7) is the soil moisture change model,

[0106] W t =W t-1 +P 0,t +K t +M t -ET t (7)

[0107] Where: t is the time; W t is the soil water content at time t; W t-1 is the soil moisture content at the previous moment; P 0,t is the effective rainfall in this period; K t is the shallow groundwater recharge during this period; M t is the irrigation amount during this period; ET t is the soil water evaporation during this period;

[0108] Formulas (8)-(11) are crop yield models.

[0109] ET A =K s ×ET(8)

[0110] Where: ET A K is the crop evapotranspiration; s is the soil moisture correction factor;

[0111]

[0112] Where: θ is the average soil moisture content in the crop root activity layer during the calculation period; θ f is the upper limit of the optimal soil moisture; θ wp is the soil wilting coefficient;

[0113]

[0114] Where: Y a , Y max , E.T. A,k , E.T. k They represent the actual crop yield, the maximum yield, the sum of the actual transpiration in production cycle k, and the sum of the actual daily crop water demand in production cycle k; B k is the moisture sensitivity empirical coefficient;

[0115]

[0116] Where: D k,j is the potential yield reduction rate of each farmland per day; j is the day sequence, Y a,k,j , Y max,k,j , E.T. A,k,j They represent the equivalent daily output value of farmland, the maximum potential daily output value, and the evaporation and transpiration of crops per day respectively;

[0117] Formula (12) is the irrigation water-to-electricity conversion model,

[0118]

[0119] Where: is the irrigation water volume of farmland i during this period; is the electricity consumption for irrigation of farmland i during this period, T C is the water-to-electricity conversion coefficient.

[0120] S3: constructing a farm virtual power plant operation cost calculation model, which is used to calculate the total cost of the farm virtual power plant's electricity consumption throughout the day and the potential production reduction. The farm virtual power plant operation cost calculation model includes constructing a virtual power plant flexibility resource model and a virtual power plant agriculture-power operation cost model;

[0121] The virtual power plant flexibility resource model includes a wind and solar power model, an energy storage device model, a reactive power compensation SVG device model, and an irrigation load model. It models the operation principle of all flexibility resources in the virtual power plant, builds the physical basis of the virtual power plant control strategy, and uses power balance constraints to ensure the safety of virtual power plant operation.

[0122] The virtual power plant agriculture-electricity operation cost model calculates the virtual power plant electricity cost, agricultural production reduction cost, and virtual power plant electricity sales revenue, and then comprehensively calculates the total benefit of the virtual power plant operation. Formula (13) is the wind and solar power model,

[0123]

[0124] Where: are the maximum output power of wind and solar power sources at time t, and the upper limits of wind and solar power sources are the predicted output values ​​after uncertainty processing; P t WT , P t PV are wind power / photovoltaic output values ​​respectively;

[0125] Formulas (14)-(16) are energy storage device models.

[0126]

[0127] Where: is the remaining energy of the battery; Divided into the initial, minimum and maximum remaining energy of the battery; It is the charge / discharge mark of the battery;

[0128]

[0129] Where: P t ESS,in , P t ESS,out are the battery charging / discharging power respectively; They are the upper limits of battery charging / discharging power respectively;

[0130]

[0131] Where: η in , η out , γ ESS are the charging / discharging efficiency / dissipation rate of the battery respectively; T is the total number of moments; ΔT is the total time period;

[0132] Formula (17) is the reactive power compensation SVG equipment model,

[0133]

[0134] Where: is the maximum reactive power compensation power of the SVG device; Q t SVG Contribute to SVG devices;

[0135] Formula (18) is the irrigation load model,

[0136]

[0137] Where: is the maximum power of the irrigation load; P t well is the operating power of the irrigation load;

[0138] Formula (19) is the power balance constraint,

[0139]

[0140] Where: Purchase electricity for virtual power plants; Purchase electricity for the virtual power plant; P fix is a fixed active load; is the reactive power balance of the power grid; Q fix is a fixed reactive load;

[0141] Formula (20) is the soil moisture constraint for agricultural production;

[0142] W i min ≤W i t ≤W i max (20)

[0143] Where: W i t is the soil moisture of farmland i at time t; W i min , W i max are the lower and upper limits of soil moisture, respectively;

[0144] The virtual power plant agriculture-electricity operation cost model is expressed by formula (21):

[0145]

[0146] Where: C is the total cost; The price of electricity purchased from the grid; The price of electricity sold to the grid; Purchase electricity for virtual power plants; Purchase electricity for virtual power plants; D i Y is the potential daily yield reduction rate in this growth stage; m i is the estimated total output value of a single farmland; N is the total number of farmlands.

[0147] In addition, this embodiment provides an optimization scheduling method for the smart farm virtual power plant model with agricultural irrigation load as described above, as follows: first, according to the typical scenario generation technology based on the neighbor propagation clustering algorithm, the typical scenario for the day-ahead scheduling reference of the virtual power plant is obtained, and the water demand and soil moisture requirements of the crops on that day are calculated according to the parameters in the typical prediction scenario; then, an agricultural-electricity coupling relationship model and a farm virtual power plant operation cost calculation model are constructed, the topological structure of the virtual power plant operation is built, and the power balance constraints and agricultural production soil moisture constraints of the virtual power plant operation are constructed; finally, the total cost model for the regulation of the flexible resources of the smart farm virtual power plant is used to run the random optimization strategy according to the obtained typical scenario, and the overall process of solving the optimization problem is as follows: Figure 2 Calculate the optimal control and dispatch scheme of the flexibility resources of the smart farm virtual power plant and obtain the day-ahead dispatch strategy;

[0148] Among them, formula (22) shows the total cost model of regulating the flexibility resources of the smart farm virtual power plant, including the electricity cost of the virtual power plant in a typical scenario, the cost of agricultural production reduction in a typical scenario, and the electricity sales income of the virtual power plant in a typical scenario.

[0149]

[0150] Where: U is the number of typical scenes after clustering; D u i is the potential daily reduction rate of the typical scenario in this growth stage, Y m,u i is the estimated total output value of a single farmland under a typical scenario; p i (u) is the probability of occurrence of a typical scenario; They are the electricity purchased and sold at time t, are the electricity purchase price and electricity sales price at time t respectively.

[0151] Example 2

[0152] The present invention can realize the modeling and optimal scheduling of the virtual power plant of the smart farm with agricultural irrigation load. The production season operation simulation case is constructed with reference to the actual scene. This case has a total of 9 farmlands, each of which covers an area of ​​300,000 mu and is planted with rice; each farmland is equipped with a 5MW smart irrigation well group, and a unified and orderly irrigation plan is formulated; in addition to the smart irrigation load, the distribution network is equipped with a 1MW wind turbine, a 1MW photovoltaic cell, and a battery with a maximum capacity of 5MW. The irrigation plan before optimization is Figure 3 The optimized irrigation plan is Figure 4Indicates. From the results, it can be seen that the No. 6-9 well group chooses short, large-scale irrigation and staggers the peaks. Most of them choose to start up during the peak period of new energy or the low period of electricity prices; No. 9 has a short and large-scale irrigation in addition to the early morning during the photovoltaic peak period. The well groups at other nodes mostly choose to irrigate multiple times and in small quantities throughout the day. The irrigation well group can implement effective staggered irrigation. They all choose to start up during the peak output of wind turbines, the peak output of photovoltaics, or the period with lower electricity prices, and stagger their respective irrigation peaks; the irrigation plan is distributed throughout the 24 hours of the day, and the concentrated power congestion caused by disordered irrigation at individual times is alleviated, which proves the effectiveness of the proposed smart farm virtual power plant modeling and optimization scheduling method.

[0153] The comparison of optimization benefits is shown in Tables 1 and 2. It can be seen that the total electricity cost of the distribution network reached 26,886.13 yuan throughout the day, and the voltage drop problem was serious, with high risk of safe operation. The electricity cost of the orderly irrigation plan was 25,978.05 yuan, which was 3.38% lower than that of the centralized disorder. The irrigation strategy can complete the irrigation task in a more economical way. At the same time, the irrigation well group can successfully complete the task of alleviating voltage congestion. The irrigation plan is distributed throughout the day for 24 hours. The concentrated power congestion phenomenon caused by disordered irrigation at individual moments is alleviated, effectively avoiding the low voltage problem caused by disordered power consumption of irrigation wells. Taking the voltage of the end node of the line where the irrigation well group is installed as an example, compared with disordered irrigation, the average voltage drop of the 18th end node is reduced by 14.22%, the maximum voltage drop is reduced by 52.83%, the average voltage drop of the 33rd end node is reduced by 13.90%, and the maximum voltage drop is reduced by 50.32%. The network security constraints can be well maintained.

[0154] Table 1 Economic analysis of irrigation strategies

[0155]

[0156] Table 2 Safety analysis of irrigation strategies

[0157]

[0158] In order to make the comparison effect of the example more obvious, the requirements for soil water layer maintenance are reduced in the comparison example, and the growth stages such as tillering period after the end of spring irrigation of rice are simulated to create a potential yield deficit scenario. The optimization strategy effect of the proposed smart farm virtual power plant modeling and optimization scheduling method is compared with the irrigation strategy effect obtained by the deterministic optimization method and the full-scenario optimization method, and the results are shown in Table 3.

[0159] Table 3 Comparison of optimization strategy efficiency

[0160]

[0161]

[0162] It can be seen from the table that if precipitation uncertainty is ignored and the irrigation plan is made directly based on the weather forecast results, the distribution network optimization scheduling scheme will be obtained at a faster speed, but the scheme does not reserve enough irrigation water margin for potential drought, resulting in nearly half of the production reduction scenarios; if the full-scenario optimization method is adopted, optimization can be performed according to all expected scenarios to ensure that there is no production reduction, but the calculation speed is extremely low and the efficiency is very poor. Compared with the deterministic optimization scheduling model, the proposed smart farm virtual power plant modeling and optimization scheduling method can greatly reduce the agricultural production damage scenario from 47% to only 3%, significantly improving the system's grain production and supply capacity; compared with the full-scenario distribution network optimization scheduling model, the proposed smart farm virtual power plant modeling and optimization scheduling method significantly reduces the calculation speed at the cost of a slight increase in the production reduction rate, which is only 0.4% of the full-scenario optimization, and the electricity cost is also reduced by about 4.2% compared with the full-scenario optimization method. While greatly improving the optimization calculation efficiency, it basically ensures the effectiveness of the optimization and ensures the reliable supply of grain production.

Claims

1. A method for constructing a smart farm virtual power plant model containing agricultural irrigation load, characterized in that: The steps include: S1: Construct a typical scenario based on the nearest neighbor propagation clustering algorithm: collect weather and electricity price forecast information of the smart farm virtual power plant, and use the clustering algorithm to obtain the typical scenario for the day-ahead dispatch of the virtual power plant, including the output information of wind power equipment, the output information of photovoltaic equipment, the fluctuation information of market electricity prices and the fluctuation information of precipitation; S2: Constructing the agriculture-electricity coupling relationship model: including constructing the crop water demand model, soil moisture change model, crop yield model and irrigation water-electricity conversion model. The crop water requirement model includes crop coefficient, crop real water requirement parameter, reference crop water requirement parameter, saturated water vapor pressure curve slope parameter, surface net radiation parameter, soil heat flux parameter, psychrometric constant, daily average temperature parameter, wind speed parameter at 2 meters above the surface, saturated water vapor pressure parameter, and actual water vapor pressure parameter; The soil moisture variation model includes soil moisture content parameters, effective rainfall parameters, shallow groundwater daily recharge parameters, irrigation parameters, and soil moisture evaporation parameters. The crop yield model includes crop evapotranspiration parameters, soil moisture correction coefficient, average soil moisture content parameters in the crop root activity layer, soil wilting coefficient, and crop water production function. The irrigation water-to-electricity conversion model includes a parameter of farmland irrigation electricity consumption multiplied by a water-to-electricity conversion coefficient; S3: constructing a farm virtual power plant operation cost calculation model, which is used to calculate the total cost of the farm virtual power plant's electricity consumption throughout the day and the potential production reduction. The farm virtual power plant operation cost calculation model includes constructing a virtual power plant flexibility resource model and a virtual power plant agriculture-power operation cost model; The virtual power plant flexibility resource model includes a wind and solar power model, an energy storage device model, a reactive power compensation SVG device model, and an irrigation load model. It models the operation principle of all flexibility resources in the virtual power plant, builds the physical basis of the virtual power plant control strategy, and uses power balance constraints to ensure the safety of virtual power plant operation. The virtual power plant agriculture-electricity operation cost model calculates the virtual power plant electricity cost, agricultural production reduction cost, and virtual power plant electricity sales revenue, and then comprehensively calculates the total benefit of the virtual power plant operation.

2. The method for constructing a smart farm virtual power plant model containing agricultural irrigation load according to claim 1 is characterized in that: In step S1, a clustering algorithm is used to obtain a typical scenario for the day-ahead dispatch of a virtual power plant, as shown in formulas (1)-(4): Formula (1) is the method for calculating the similarity matrix elements; Where: s(m,n) is the similarity matrix element, the non-diagonal elements are Euclidean distances; m, m', n, n' are the point numbers; x m 、x n are the uncertain parameters respectively; Formula (2) is the method for calculating the elements of the attraction matrix; In the formula: r(m,n) and a(m,n') are the elements of the attraction matrix and the belonging matrix respectively; Formula (3) is the calculation method of the attribution matrix elements; Formula (4) attenuation iterative update calculation; Where λ represents the damping coefficient and s represents the number of iterations.

3. The method for constructing a smart farm virtual power plant model containing agricultural irrigation load according to claim 2 is characterized in that: In step S2, formulas (5)-(6) are crop water requirement models. ET=K C ×AND R (5) Where: K C is the crop coefficient, ET is the actual water requirement of the crop, and ET R is the reference crop water requirement; Where: Δ is the slope of the saturated water vapor pressure curve; R n is the net radiation of the surface; G is the soil heat flux; γ is the psychrometric constant; T mean is the daily average temperature; u2 is the wind speed at 2 meters above the ground; e s is the saturated water vapor pressure; e a is the actual water vapor pressure; Formula (7) is the soil moisture change model, W t =W t-1 +P 0,t +K t +M t -AND t (7) Where: t is the time; W t is the soil water content at time t; W t-1 is the soil moisture content at the previous moment; P 0,t is the effective rainfall in this period; K t is the shallow groundwater recharge during this period; M t is the irrigation amount during this period; ET t is the soil water evaporation during this period; Formulas (8)-(11) are crop yield models. AND A =K s ×AND (8) Where: ET A K is the crop evapotranspiration; s is the soil moisture correction factor; Where: θ is the average soil moisture content in the crop root activity layer during the calculation period; θ f It is the upper limit of the optimum soil moisture; θ wp is the soil wilting coefficient; Where: Y a , Y max , E.T. A,k , E.T. k They represent the actual crop yield, the maximum yield, the sum of the actual transpiration in production cycle k, and the sum of the actual daily crop water demand in production cycle k; B k is the moisture sensitivity empirical coefficient; Where: D k,j is the potential yield reduction rate of each farmland per day; j is the day sequence, Y a,k,j , Y max,k,j , E.T. A,k,j They represent the equivalent output value of farmland per day, the maximum potential output value per day, and the evaporation and transpiration of crops per day respectively; Formula (12) is the irrigation water-to-electricity conversion model, Where: is the irrigation water volume of farmland i during this period; is the electricity consumption for irrigation of farmland i during this period, T C is the water-to-electricity conversion coefficient.

4. The method for constructing a smart farm virtual power plant model containing agricultural irrigation load according to claim 3 is characterized in that: In step S3, formula (13) is the wind and solar power model, Where: are the maximum output power of wind and solar power sources at time t, respectively. The upper limits of wind and solar power sources are the predicted output values ​​after uncertainty processing; are wind power / photovoltaic output values ​​respectively; Formulas (14)-(16) are energy storage device models. Where: is the remaining energy of the battery; Divided into the initial, minimum and maximum remaining energy of the battery; It is the charge / discharge mark of the battery; Where: are the battery charging / discharging power respectively; They are the upper limits of battery charging / discharging power respectively; Where: η in , η out , γ ESS are the charging / discharging efficiency / dissipation rate of the battery respectively; T is the total number of moments; ΔT is the total time period; Formula (17) is the reactive power compensation SVG equipment model, Where: is the maximum reactive compensation power of the SVG device; Contribute to SVG devices; Formula (18) is the irrigation load model, Where: is the maximum power of the irrigation load; is the operating power of the irrigation load; Formula (19) is the power balance constraint, Where: Purchase electricity for virtual power plants; Purchase electricity for the virtual power plant; P fix is a fixed active load; is the reactive power balance of the power grid; Q fix is a fixed reactive load; Formula (20) is the soil moisture constraint for agricultural production; Where: is the soil moisture of farmland i at time t; are the lower and upper limits of soil moisture respectively; the agriculture-electricity operation cost model of the virtual power plant is expressed by formula (21): Where: C is the total cost; The price of electricity purchased from the grid; The price of electricity sold to the grid; Purchase electricity for virtual power plants; Purchase electricity for virtual power plants; D i Y is the potential daily yield reduction rate in this growth stage; m i is the estimated total output value of a single farmland; N is the total number of farmlands.

5. A smart farm virtual power plant model system including agricultural irrigation load, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements a method for constructing a virtual power plant model of an intelligent farm containing agricultural irrigation loads as described in any one of claims 1 to 4.

6. The optimization scheduling method of a smart farm virtual power plant model system containing agricultural irrigation load according to claim 5 is characterized in that: The method is as follows: first, according to the typical scenario generation technology based on the nearest neighbor propagation clustering algorithm, the typical scenario for the reference of the day-ahead dispatch of the virtual power plant is obtained, and the water demand and soil moisture requirements of the crops on the day are calculated according to the parameters in the typical prediction scenario; then, the agricultural-electricity coupling relationship model and the farm virtual power plant operation cost calculation model are constructed, the topological structure of the virtual power plant operation is built, and the power balance constraints and agricultural production soil moisture constraints of the virtual power plant operation are constructed; finally, the total cost model of the regulation of the flexible resources of the smart farm virtual power plant is used, and the random optimization strategy is run according to the obtained typical scenario to calculate the optimal regulation and dispatching scheme of the flexible resources of the smart farm virtual power plant, and obtain the day-ahead dispatching strategy; Among them, formula (22) shows the total cost model of regulating the flexibility resources of the smart farm virtual power plant, including the electricity cost of the virtual power plant in a typical scenario, the cost of agricultural production reduction in a typical scenario, and the electricity sales income of the virtual power plant in a typical scenario. Where: U is the number of typical scenes after clustering; D u i is the potential daily reduction rate of the typical scenario in this growth stage, Y m,u i is the estimated total output value of a single farmland under a typical scenario; p i (u) is the probability of occurrence of a typical scenario; They are the electricity purchased and sold at time t, are the electricity purchase price and electricity sales price at time t respectively.

7. An optimization dispatching system for a smart farm virtual power plant with agricultural irrigation load, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program implements the optimization scheduling method as described in claim 6 when executed by the processor.

Images