A method for optimizing configuration of a fire-storage combined frequency modulation system

Abstract

The application discloses a kind of fire storage combined frequency modulation system optimization configuration methods, first establish the total cost loss model of fire storage combined frequency modulation system, the total economic benefit model of fire storage combined frequency modulation system is obtained by subtracting the total cost loss model from the frequency modulation compensation benefit model of grid side;Secondly based on the day-ahead load curve of thermal power generating unit issued by grid side, the optimal output of thermal power generating unit and energy storage battery is solved using artificial bee colony optimization algorithm and the trend chart of energy storage state of charge is obtained;Finally, the optimized output value is applied to the fire storage combined frequency modulation simulation response process, and the state of charge of energy storage is adjusted based on the above trend chart in advance. Compared with the fire storage combined frequency modulation strategy considering only frequency modulation benefit, the total economic benefit of fire storage combined frequency modulation system during frequency modulation is improved under the method, and the availability of energy storage battery is further improved.

Description

TECHNICAL FIELD

[0001] The present application relates to a kind of fire storage combined frequency modulation system optimization configuration technology, belongs to fire storage combined frequency modulation system load optimization distribution strategy technical field. BACKGROUND

[0002] The current fire storage combined frequency modulation load optimization distribution method mainly focuses on the maximization of power grid frequency modulation compensation benefit, lacks the detailed consideration of the cost loss of thermal power unit and energy storage battery in the process of fire storage combined frequency modulation, and the total economic benefit of fire storage combined frequency modulation system does not reach the theoretical optimum. In the actual frequency modulation instruction response process, thermal power unit often responds to the frequency modulation instruction with the maximum regulation rate it can reach, which will increase the frequency modulation cost of thermal power unit in all aspects, such as the increase of total coal consumption and the increase of environmental protection payment cost. In addition, energy storage only makes up the power difference, and the frequent charging and discharging capacity leads to overcharge or overdischarge of battery, and there is lack of more target-oriented state of charge adjustment method. SUMMARY

[0003] Technical problem: the present application aims to solve the technical problem that the existing fire storage combined frequency modulation instruction response method does not fully consider the cost loss of fire storage combined frequency modulation and the lack of more explicit pre-charging and discharging guidance reference for state of charge adjustment of energy storage, and proposes a fire storage combined frequency modulation system optimization configuration method. Under the application of the method, the power output of fire storage combined frequency modulation not only considers the frequency modulation compensation benefit, but also considers the cost loss of fire storage combined frequency modulation. In addition, the state of charge of energy storage will deviate from the upper and lower limits of 0.2 or 0.8, and the availability will be further improved.

[0004] Technical scheme: the fire storage combined frequency modulation system optimization configuration method of the present application comprises the following steps:

[0005] S1: a total cost loss mathematical model of fire storage combined frequency modulation system is established;

[0006] S2: the total cost loss mathematical model is subtracted from the power grid frequency modulation compensation benefit model to obtain a total economic benefit model of fire storage combined frequency modulation system frequency modulation;

[0007] S3: select the load curve of thermal power unit issued in advance by power grid as the equality constraint condition to be optimized, and perform optimization calculation under the inequality constraint condition of meeting the basic power upper and lower limits of thermal power unit and energy storage battery to obtain the fire storage optimization output value

[0008] S4: the artificial bee colony optimization algorithm is used to calculate the optimization output of thermal power unit and energy storage and obtain the trend chart of state of charge of energy storage;

[0009] S5: the fire storage optimization output value obtained by optimization calculation is applied to the variable load process of fire storage combined frequency modulation, so that the total economic benefit of fire storage combined frequency modulation system frequency modulation tends to be theoretically optimal.

[0010] S6: According to the obtained energy storage state of charge change trend diagram, in the frequency modulation waiting instruction process, the fluctuating power of the thermal power unit is used to pre-adjust the state of charge of the energy storage battery, so that the state of charge SOC value of the energy storage battery is away from the upper and lower limit requirements of 0.2 or 0.8, and the availability index is improved.

[0011] The step S1 comprises: modeling the cost loss of the thermal power unit and the energy storage battery in the frequency modulation process respectively; first, modeling the frequency modulation cost of the thermal power unit, considering the modeling of the coal consumption cost of the thermal power unit frequency modulation and the modeling of the environmental payment cost, the modeling of the coal consumption cost of the thermal power unit frequency modulation is subdivided into the modeling of the steady-state coal consumption and the modeling of the transient-state coal consumption, and the modeling of the environmental payment cost of the thermal power unit is the modeling of the environmental payment cost caused by the pollution emission of the unit; second, modeling the cost loss of the energy storage battery in the frequency modulation process as the life loss cost model caused by its discharge; finally, adding the life loss cost model of the energy storage battery in the frequency modulation process to the coal consumption cost model and the environmental payment cost model of the thermal power unit frequency modulation to obtain the total cost loss mathematical model of the thermal storage combined frequency modulation system.

[0012] The total cost loss mathematical model of the thermal storage combined frequency modulation system is:

[0013] C zong =C mei +C pai +C batlife

[0014] Wherein: C zong is the total loss cost of the thermal storage combined frequency modulation process; C mei is the total coal consumption cost of the thermal storage combined frequency modulation process; C pai is the environmental payment cost of the thermal storage combined frequency modulation process; C batlife is the battery life loss cost of the thermal storage combined frequency modulation process;

[0015] The coal consumption cost fitting model of the thermal power unit is as follows:

[0016]

[0017] Wherein: C mei is the coal consumption cost per unit hour; V e is the load regulation rate of the thermal power unit; C price is the real-time coal price, m wen is the power supply coal consumption of the thermal power unit under steady state, P is the unit load, a1, a2, a3, a4 are 0.003, 0.013, 0.001, -0.176 respectively;

[0018] The environmental payment cost model is as follows:

[0019]

[0020] wherein: λ is an adjustment factor; P NOx , P SO2 are the emission values of NOx and SO2 respectively; N1 and N2 are the equivalent values of NOx and SO2 emission values respectively;

[0021] The life loss cost model of energy storage is as follows:

[0022]

[0023] C ini is the initial investment cost of the battery; d is the single discharge ampere-hour number of each variable load process; the total effective discharge amount in the life cycle can be represented by the following formula:

[0024] Γ R = L R D R C R

[0025] wherein: L R is the rated cycle life of the battery; D R is the rated discharge depth; C R is the rated capacity of the battery.

[0026] The steady-state coal consumption modeling involves the coal consumption of a unit of power generation when the thermal power generating unit is at a fixed power output, and the fitting and regression analysis of coal consumption data of various types of main units to obtain the change model of the power of the thermal power generating unit and the coal consumption; the basic mathematical model thereof is as follows:

[0027]

[0028] wherein: m wen is the steady-state power supply coal consumption of the thermal power generating unit, P is the unit load; m0 is the reference coal consumption, P0 is the reference load; a = 0.798; b = 0.988; u0 = 0.0035; t = 0.245; β0 = -0.0348.

[0029] In the step S2, the total cost loss mathematical model C zong is subtracted from the grid frequency modulation compensation benefit model Gain to obtain the total economic benefit model OBJ of the combined frequency modulation system of the fire and storage,

[0030] OBJ = C zong -Gain

[0031] The grid frequency modulation compensation benefit model Gain is:

[0032] Gain = KDQ

[0033] Wherein: K is the frequency modulation comprehensive performance index, D is the frequency modulation mileage, generally refers to the absolute value of the difference between the actual output and the initial output after the unit responds to the instruction, and Q is the frequency modulation compensation price.

[0034] In the step S3, the load curve of the thermal power unit in advance issued by the power grid is selected as an equal constraint condition to be satisfied in optimization, and optimization calculation is performed under inequality constraint conditions conforming to the basic power upper and lower limits of the thermal power unit and the energy storage battery. The thermal power unit day-ahead load curve is the next day load prediction curve issued in advance by the power grid side, the sum of the outputs of the thermal power unit and the energy storage battery should take the instruction value as the adjustment target to meet the basic instruction power requirement, and the fire storage optimization output value is obtained:

[0035]

[0036] The above formula shows that the optimization output value of the thermal power unit should be obtained between the minimum allowable value and the maximum allowable value;

[0037]

[0038] The above formula shows that the optimization output value of the energy storage battery should be obtained between the minimum allowable value and the maximum allowable value.

[0039] In the step S4, the artificial bee optimization algorithm is used to calculate the optimization output of the thermal power unit and the energy storage and obtain the change trend graph of the state of charge of the energy storage. The algorithm logic sequence of the artificial bee optimization algorithm under the model of the method is as follows: the first step is initialization, the basic parameters of the thermal storage device such as power limit are input, then the artificial bee optimization algorithm parameters such as determining the number of bees as 200, the maximum number of iterations as 300, and generating the initial random solution x i , i = 1, 2, …, 100; in the second step, the employed bees generate new random solutions s ij in the neighborhood of the initial random solution according to the following formula: The superiority of the new solution and the initial solution is calculated and compared, and the solution with better fitness is left;

[0040] s ij = x ij + β ij (x ij -x kj )

[0041] Wherein: k ∈ {1, 2, …, 100}, k ≠ i; j ∈ {1, 2}; β ij is a random number between -1 and 1;

[0042] In the third step, the employed bees share the solution information with the followed bees, and the probability value of each solution is calculated according to the following formula:

[0043]

[0044] wherein: fit i is the fitness value of each solution, i.e. the degree of goodness or badness of the solution.

[0045] Then a random number between -1 and 1 is generated based on the roulette method, if the random number is less than the probability value of the solution, the current optimal solution is retained according to the second step, and the step is continued;

[0046] After the bee search in the fourth step is completed, it is checked whether there is a local optimal solution, i.e. whether there is a solution that has not been updated for more than the maximum number of times M of non-updating, if there is, the local optimal solution is discarded, and a new solution is generated according to the following formula:

[0047] x ij = x mini + a (x maxj - x minj )

[0048] wherein: a is a random number between 0 and 1; x mini , x maxj , x minj are the minimum value and the maximum value of the solution corresponding to the dimension; and then the above steps are repeated.

[0049] In the fifth step, it is checked whether the maximum number of iterations 300 has been reached, if it has been reached, the optimal solution can be output.

[0050] Finally, the optimized output value of the fire storage combined frequency modulation system is obtained, and the trend chart of the state of charge of the energy storage is further obtained.

[0051] In the step S5, the fire storage optimized output value obtained by the optimization calculation is applied to the fire storage combined frequency modulation variable load process, so that the total economic benefit of the fire storage combined frequency modulation system tends to be theoretically optimal when frequency modulation is performed; when applied, the load instruction prediction value issued in advance by the power grid side needs to be compared with the actual load instruction value, i.e. when the deviation between the day-ahead prediction load instruction and the actual load instruction value is less than a certain allowed value L, the fire storage optimal output value calculated in advance is applied to the variable load process with smaller load instruction deviation, which is equivalent to performing optimization calculation in advance; when the deviation between the day-ahead prediction load instruction and the actual load instruction is greater than the allowed value L, the deviation between the result of the optimization calculation in advance and the actual fire storage optimal output value exceeds the allowed range, at this time, the frequency modulation instruction response is performed according to the current fire storage output matching mode of the corresponding power plant, i.e. at this time, the pre-calculated fire storage optimized output value cannot be used.

[0052] The step S6 is to pre-adjust the state of charge of the energy storage battery, provided that the thermal power unit and the energy storage battery have responded to the previous frequency modulation instruction, the grid-side frequency modulation performance evaluation index has been fully evaluated, and a new frequency modulation instruction has not yet appeared; at this time, the thermal power unit will make small-range power fluctuations around the required power due to inertia, and the fluctuating power is used to pre-charge and discharge the energy storage to adjust the state of charge.

[0053] The pre-charge and discharge is to refer to the state of charge trend graph of the energy storage battery obtained by prior optimization calculation, and the direction and size of the state of charge change at the next time point in the trend graph are used to adjust the current state of charge, and there is a time point every 5 minutes, and the adjustment target is to make the state of charge of the energy storage far away from the upper and lower limits of 0.2 or 0.8, and improve the availability of the energy storage.

[0054] Advantages: Compared with the prior art, the present application has the following advantages:

[0055] (1) The present application can make the thermal power unit and the energy storage battery tend to the theoretical optimal economic benefit of the combined frequency modulation system of the thermal power unit and the energy storage battery when responding to the grid frequency modulation instruction, and in a long time scale, the optimized output of the combined frequency modulation system of the thermal power unit and the energy storage battery can bring comprehensive benefit improvement to the power plant side, mainly reducing the frequency modulation loss cost of the power plant side.

[0056] (2) When pre-charging and discharging the energy storage battery, the state of charge trend graph obtained by optimization calculation can be used to adjust the power in a more directional manner, so that the availability of the energy storage battery is further improved. BRIEF DESCRIPTION OF DRAWINGS

[0057] Figure 1 is the flow framework diagram of the method of the present application;

[0058] Figure 2 is a structural schematic diagram of the combined frequency modulation system of the thermal power unit and the energy storage battery;

[0059] Figure 3 is a daily load curve diagram of the thermal power unit issued in advance by the grid on a certain day;

[0060] Figure 4 is a flow diagram of the optimization calculation method in the present application;

[0061] Figure 5 is a schematic diagram of the optimized output of the combined frequency modulation system of the thermal power unit and the energy storage battery in the example;

[0062] Figure 6 is a state of charge trend graph of the energy storage battery in the example. DETAILED DESCRIPTION

[0063] The technical solution will be described in detail below in combination with the simulation example and the drawings:

[0064] As Figure 1 shown, an optimal configuration method of a fire storage combined frequency modulation system considering economic benefits, comprising:

[0065] S1: Establish a total cost loss mathematical model of a fire storage combined frequency modulation system. First, establish a coal consumption cost loss model of a thermal power unit. Here, it is divided into steady-state coal consumption modeling and transient coal consumption modeling. Steady-state coal consumption modeling involves the amount of coal consumed per unit of power generation when the thermal power unit is at a fixed power output. The coal consumption data of various types of main units in China can be fitted and regression analyzed to obtain a model of the change of thermal power unit power and coal consumption. The basic mathematical model is as follows:

[0066]

[0067] Wherein: m wen is the steady-state power supply coal consumption of the thermal power unit, g / (kW h); P is the unit load, MW; m0 is the reference coal consumption, g / (kWh); P0 is the reference load, MW; and the rest are fitting coefficients.

[0068] Transient coal consumption mainly reflects the transient coal consumption increment caused by the variable load of the thermal power unit, which is related to the frequency modulation rate of the thermal power unit. A relevant dynamic model data can be referred to for modeling. Finally, the coal consumption cost fitting model of the thermal power unit is as follows:

[0069]

[0070] Wherein: C mei is the coal consumption cost per unit hour; V e is the load adjustment rate of the thermal power unit, MW / min; C price is the real-time coal price.

[0071] Secondly, the environmental protection payment cost of the thermal power unit during frequency modulation is also considered in the optimization objective function. For the unit pollutant emission cost, mainly considering NOx, SO2, as shown in the following formula:

[0072]

[0073] Wherein: λ is the adjustment coefficient; P NOx , P SO2 are the emission values of NOx and SO2 respectively; N1 and N2 are the equivalent values of NOx and SO2 emission values respectively.

[0074] The unit pollutant emission value is closely related to the working efficiency of the environmental protection equipment. Here, a fitting model of the unit working condition and the removal efficiency of the environmental protection equipment is constructed by referring to the dynamic model data. Taking the pollutant NOx as an example, there is the following emission calculation model:

[0075]

[0076] wherein: b tp is the coal consumption for power generation, g / (kWh); N is the percentage of nitrogen contained in coal; P is the unit operating value, MW; η m is the conversion rate of fuel nitrogen; n is 80%; m1=1x10 -8 , m2=1.2x10 -6 , m3=6x10 -4 , m4=0.082.

[0077] Further considering the frequency regulation life loss cost of the energy storage battery, mainly considering the discharge life loss cost. The total effective discharge capacity in its life cycle can be represented by the following formula:

[0078] Gamma R = L R D R C R

[0079] wherein: L R is the rated cycle life of the battery; D R is the rated discharge depth; C R is the rated capacity of the battery.

[0080] The life loss cost is estimated by the ratio of each discharge capacity to the total effective discharge capacity, as follows:

[0081]

[0082] wherein: C ini is the initial investment cost of the battery; d is the single discharge ampere-hour of each variable load process.

[0083] Finally, the above cost categories are superimposed to obtain the total cost loss model of the fire storage combined frequency regulation as follows:

[0084] C zong = C mei + C pai + C batlife

[0085] S2: The total cost loss mathematical model is subtracted from the grid frequency regulation compensation benefit model to obtain the total economic benefit model of the fire storage combined frequency regulation system, which enriches the optimization objective function. The grid side will check and compensate the comprehensive condition of the fire storage combined frequency regulation system, which can be described by the following formula:

[0086] Gain = KDQ

[0087] Wherein: K is the frequency modulation comprehensive performance index, which can be calculated by referring to the compensation rules of the corresponding power grid; D is the frequency modulation mileage, generally refers to the absolute value of the difference between the actual output and the initial output of the unit after responding to the instruction, MW; Q is the frequency modulation compensation price, which is set to 5-10 yuan / MW here.

[0088] In this way, the total economic benefit model of the fire storage combined frequency modulation system is obtained by subtracting the benefit from the cost, which is also the total objective function in the optimization calculation as follows:

[0089] OBJ = C zong -Gain

[0090] S3: Select the power grid issued in advance thermal power unit load curve as the optimization of the equation constraint conditions, while meeting the basic power upper and lower limit of thermal power unit and energy storage battery inequality constraint conditions for optimization calculation. The thermal power unit day-ahead load curve here is the next day load forecast curve issued in advance by the power grid side, which has reference value, and the sum of the output of thermal power unit and energy storage battery should take the instruction value as the adjustment target to meet the basic instruction power requirement. In addition, the optimization also needs to meet the relevant inequality constraint conditions as shown below:

[0091]

[0092] The above formula shows that the optimized output value of the thermal power unit should be between the minimum allowed value and the maximum allowed value.

[0093]

[0094] The above formula shows that the optimized output value of the energy storage battery should be between the minimum allowed value and the maximum allowed value.

[0095] S4: The artificial bee colony optimization algorithm is used to calculate the optimized output of the thermal power unit and the energy storage and obtain the trend chart of the state of charge of the energy storage. The use process of the artificial bee colony optimization algorithm under the model in this paper is as shown in Figure 4 The final optimized output value of the fire storage combined frequency modulation system is calculated, and the trend chart of the state of charge of the energy storage can be further obtained.

[0096] S5: the fire storage optimization output value obtained by the optimization calculation is applied to the fire storage combined frequency modulation variable load process, so that the total economic benefit of the fire storage combined frequency modulation system tends to be theoretically optimal when frequency modulation; when applying, the load instruction prediction value issued in advance by the power grid side needs to be compared with the actual load instruction value, i.e. when the deviation between the day-ahead predicted load instruction and the actual load instruction value is less than a certain allowed value L=5MW, the pre-calculated fire storage optimal output value is applied to the variable load process with smaller load instruction deviation, which is equivalent to pre-optimization calculation. For example, the current frequency modulation instruction of the unit is P1MW, the current output is P2MW, and the pre-predicted value is (P1+2)MW. At this time, the deviation between the predicted value and the actual instruction is less than the allowed value. The pre-calculated optimal output of the unit is P3MW, and the optimal output of the energy storage battery is P4MW. The time interval is 5 minutes, so the rate allocated to the unit should be as follows:

[0097]

[0098] Where: Vubest is the optimal variable load rate allocated to the unit at this time, MW / min.

[0099] That is, when the fire storage combined frequency modulation system responds to the P1 instruction, due to the small deviation between the predicted value and the actual value, the pre-calculated optimization value is used, i.e. the response rate allocated to the unit is Vubest MW / min, and the energy storage battery assists the unit with the optimal output P4 in the instruction response process. The pre-optimization objective is to maximize the total economic benefit of the fire storage combined frequency modulation system, so the application of the optimal output at this time can make the optimization objective, i.e. the total economic benefit of the fire storage combined frequency modulation system, tend to be maximum.

[0100] In addition, when the deviation between the day-ahead predicted load instruction and the actual load instruction is greater than the allowed value L, the deviation between the pre-optimization calculation result and the actual fire storage optimal output value exceeds the allowed range, at which time the fire storage combined frequency modulation system can respond to the frequency modulation instruction according to the existing fire storage output coordination mode of the corresponding power plant, i.e. the pre-calculated fire storage optimal output value cannot be used at this time.

[0101] S6: pre-adjust the state of charge of the energy storage battery, provided that the thermal power unit and the energy storage battery have responded to the previous frequency modulation instruction, the frequency modulation performance evaluation index of the power grid side has been fully evaluated, and a new frequency modulation instruction has not yet appeared; at this time, the thermal power unit will make small-range power fluctuations around the required power due to inertia, and the pre-charge and discharge of the energy storage, i.e. the adjustment of the state of charge, is utilized.

[0102] The pre-charge and discharge is based on the state of charge trend chart obtained by prior optimization calculation, and the current state of charge is adjusted according to the direction and size of the state of charge change at the next time point in the trend chart. There is one time point every 5 minutes, and the adjustment target is to make the state of charge of the energy storage far away from the upper and lower limits of 0.2 or 0.8, so as to improve the availability of the energy storage.

[0103] The steps S5 and S6 are described below in combination with simulation example results:

[0104] Taking the load curve of a thermal power unit received by a power plant on a certain day as an example of the equation constraint, the power of the thermal power unit is 330 MW, the maximum charge and discharge power of the energy storage battery is 9 MW, and the capacity is 4.5 MWh. According to the calculation flow of the optimization algorithm shown in Figure 4 , the fire storage joint frequency modulation optimization output value is calculated, and the result is shown in Figure 5 . There is one optimization calculation point every 5 minutes, and the calculation result is the fire storage joint frequency modulation optimization output value under the constraint required by the instruction, which makes the total economic benefit of the fire storage joint frequency modulation system tend to be optimal, and the trend chart of the state of charge of the energy storage battery under the response of the optimization output is further drawn, as shown in Figure 6 .

[0105] The above is the optimization calculation process of the fire storage joint frequency modulation system. The following further describes the method of adjusting the state of charge of the energy storage in advance: Figure 6 The optimization calculation time interval is 5 minutes, and the actual frequency modulation instruction does not appear regularly, so the optimization output taken by each new frequency modulation instruction is the optimization output value at the time point one time interval before the appearance time of the instruction. The pre-adjustment of the state of charge of the energy storage is based on the trend value at the next time point, i.e. the trend value at the next 5-minute time point. When the current frequency modulation instruction has been responded to, the grid side frequency modulation evaluation index has been evaluated, and a new frequency modulation instruction has not appeared, the pre-adjustment time of the state of charge of the energy storage is entered, at which time the state of charge value at the closest next 5-minute time point on the timeline is found according to Figure 6 , and the pre-charge and discharge is performed according to the change direction.

[0106] If the state of charge value at the next 5-minute time point Figure 6 will increase, the energy storage is pre-discharged using the unit fluctuation power while waiting for the new frequency modulation instruction, i.e. a certain amount of electricity is discharged in advance to make the state of charge of the energy storage far away from the upper limit of 0.8. If the state of charge value decreases, the method is similar. In addition, the length of time of the frequency modulation waiting instruction is not fixed, so the upper and lower limits of 0.4 or 0.6 for pre-charge and discharge are set to limit the pre-charge and discharge. If the pre-charge and discharge has made the state of charge value of the energy storage reach the limit of 0.4 or 0.6, the pre-charge and discharge is immediately stopped. If a new frequency modulation instruction has appeared, the pre-charge and discharge of the energy storage is also immediately stopped, and a new round of optimization output response process is entered.

[0107] For S5, S6: A constant rate response method can be set as a control in which the thermal power unit always selects the maximum adjustment rate it can achieve to respond, and the energy storage makes up the power difference. Select a certain 8-hour frequency modulation instruction historical data segment, compare the pros and cons of the simulation results by calculating the corresponding cost and frequency modulation income items. As shown in the following table:

[0108] Table 1 Comparison of 8-hour simulation results of constant rate response method and optimized rate response method in this paper

[0109]

[0110]

[0111] From the simulation results in the above table: With the same 8-hour frequency modulation instruction historical data, but using different fire storage response methods will affect the frequency modulation cost in each dimension and the total economic benefit. After using the optimized output to respond, the total economic benefit is improved to a certain extent, mainly reflected in the reduction of total coal consumption cost, the reduction of environmental cost and the increase of frequency modulation income in table 1. Because the thermal power unit uses the optimized output to respond, in some period, the energy storage needs to discharge more to respond to the requirements of frequency modulation instruction, the increase of discharge Ah number leads to the certain growth of the life loss cost of energy storage, but the total economic benefit is better than that of constant rate response method. In addition, due to the addition of state of charge pre-adjustment of energy storage in the waiting period of frequency modulation instruction, the state of charge of energy storage battery will be as far as possible away from the upper and lower limit requirements of 0.8 or 0.2, which improves its availability rate in the frequency modulation period, which also guarantees the total economic benefit of the thermal storage combined frequency modulation system to be more optimal.

Claims

1. A method for optimizing configuration of a combined heat and power frequency regulation system, characterized by, The method comprises the following steps: S1: a total cost loss mathematical model of a fire storage combined frequency modulation system is established; S2: a total economic benefit model of the fire storage combined frequency modulation system is obtained by subtracting a power grid frequency modulation compensation benefit model from the total cost loss mathematical model; S3: a load curve of a thermal power unit issued in advance by the power grid is selected as an equal constraint condition that needs to be met in optimization, and optimization calculation is performed under inequality constraint conditions that meet power upper and lower limits of the thermal power unit and the energy storage battery, so that fire storage optimization output values are obtained S4: an artificial bee colony optimization algorithm is used to calculate the fire storage optimization output of the thermal power unit and the energy storage battery and obtain a variation trend diagram of a state of charge of the energy storage battery; S5: the fire storage optimization output values obtained through optimization calculation are applied to a fire storage combined frequency modulation variable load process, so that the total economic benefit of the fire storage combined frequency modulation system tends to be theoretically optimal when frequency modulation is performed; S6: according to the obtained state of charge variation trend diagram of the energy storage battery, the state of charge of the energy storage battery is adjusted in advance by using fluctuation power of the thermal power unit during frequency modulation waiting instruction, so that the state of charge, that is, the SOC value of the energy storage battery, is far away from upper and lower limit requirements of 0.2 or 0.8, and the availability index of the energy storage battery is improved; wherein, The step S1 comprises: respectively modeling cost losses of the thermal power unit and the energy storage battery during frequency modulation; firstly, modeling the frequency modulation cost of the thermal power unit, considering modeling of a frequency modulation coal consumption cost of the thermal power unit and modeling of an environmental protection payment cost; the modeling of the frequency modulation coal consumption cost of the thermal power unit is subdivided into modeling of a steady-state coal consumption and modeling of a transient-state coal consumption, and the modeling of the environmental protection payment cost of the thermal power unit is modeling of an environmental payment cost caused by unit pollutant emission; secondly, modeling the cost loss of the energy storage battery during frequency modulation as modeling of a life loss cost caused by discharging; finally, adding the modeling of the frequency modulation coal consumption cost and the modeling of the environmental protection payment cost of the thermal power unit to the modeling of the life loss cost of the energy storage battery during frequency modulation, to obtain the total cost loss mathematical model of the fire storage combined frequency modulation system; The total cost loss mathematical model of the fire storage combined frequency modulation system is as follows: C zong = C mei + C pai + C batlife Wherein: C zong is the total loss cost of the fire storage combined frequency modulation process; C mei is the total coal consumption cost of the fire storage combined frequency modulation process; C pai is the environmental protection payment cost of the fire storage combined frequency modulation process; C batlife is the battery life loss cost of the fire storage combined frequency modulation process; The fitting model of the coal consumption cost of the thermal power unit is as follows: C mei = (a1V e 3 +a2V e 2 +a3V e +a4+m wen ) x P x C price Wherein: C mei is the coal consumption cost per unit hour; V e is the load adjustment rate of the thermal power unit; C price is the real-time coal price, m wen is the coal consumption of the thermal power unit under steady state, P is the unit load, and a1, a2, a3, and a4 are respectively 0.003, 0.013, 0.001, and -0.

176. The environmental protection payment cost model is as follows: wherein: λ is an adjustment factor; P NOx , P SO2 are the emission values of NOx, SO2, respectively; N1, N2 are the equivalent values of the NOx, SO2 emission values, respectively; The life loss cost model of the energy storage is as follows: C ini is the initial investment cost of the battery; d is the single discharge ampere-hour number of each load change process; and the total effective discharge amount in the life cycle can be expressed by the following equation: Γ R = L R D R C R wherein: L R is the battery rated cycle life; D R is the rated depth of discharge; C R is the battery rated capacity.

2. The method of claim 1, wherein the method further comprises: The steady-state coal consumption modeling relates to coal consumption of a unit power generation when the thermal power unit is at a fixed power output, and a change model of power and coal consumption of the thermal power unit is obtained through fitting and regression analysis of coal consumption data of various types of main units; the mathematical model is as follows: wherein: m wen is the coal consumption of the thermal power unit under steady state, P is the unit load; m0 is the reference coal consumption, P0 is the reference load; a = 0.798; b = 0.988; u0 = 0.0035; t = 0.245; β0 = -0.0348.

3. The method of claim 1, wherein the method further comprises: In the step S2, the total cost loss mathematical model C zong Subtracting the grid frequency modulation compensation benefit model Gain, the total economic benefit model OBJ of the combined frequency modulation system of the power grid and the thermal storage is obtained. OBJ = C zong -Gain The power grid frequency modulation compensation benefit model Gain is as follows: Gain=KDQ Wherein: K is a frequency modulation comprehensive performance index, D is a frequency modulation mileage, generally referring to an absolute value of a difference between an actual output after the unit responds to an instruction and an initial output, and Q is a frequency modulation compensation price.

4. The method of claim 1, wherein, In the step S3, the load curve of the thermal power unit issued in advance by the power grid is selected as an equal constraint condition that needs to be met in optimization, and optimization calculation is performed under inequality constraint conditions that meet power upper and lower limits of the thermal power unit and the energy storage battery, wherein the day-ahead load curve of the thermal power unit is a next-day load prediction curve issued in advance by the power grid side, the sum of the outputs of the thermal power unit and the energy storage battery should take an instruction value as an adjustment target to meet an instruction power requirement, and fire storage optimization output values are obtained: The formula shows that the optimal output value of the thermal power unit should be between its minimum and maximum allowable values; The formula shows that the optimal output value of the energy storage battery should be between its minimum and maximum allowable values.

5. The method of claim 1, wherein, In step S4, the artificial bee colony optimization algorithm is used to calculate the optimized output of the thermal power unit and the energy storage, and to obtain the trend diagram of the energy storage's state of charge. The algorithm logic sequence of the artificial bee colony optimization algorithm under this method model is as follows: The first step is to initialize, input the parameters of the thermal power unit and the energy storage equipment, such as the power limit, and then input the parameters of the artificial bee colony optimization algorithm, such as determining the number of bees to be 200 and the maximum number of iterations to be 300, and generating an initial random solution x. i In the second step, the hired bees generate new random solutions s in the neighborhood of the initial random solution using the following formula. ij The merits of the new solution and the initial solution are calculated and compared, and the solution with better fitness is retained; s ij = x ij + β ij (x ij - x kj ) wherein: k e {1,2,...,100}, k ≠ i; j e {1,2}; β ij is a random number between -1 and 1; In the third step, the employed bees share the information of the solution with the following bees, and the probability value of each solution is calculated according to the following formula: where: fit i is the fitness value corresponding to each solution, i.e. the goodness of the solution. Then, a random number between-1 and 1 is generated based on the roulette method, and if the random number is less than the probability value of the solution, the current optimal solution is retained according to the second step. In the fourth step, after the search of the following bees is completed, it is checked whether there is a local optimal solution, i.e., whether there is a solution that has not been updated for more than the maximum number of non-updates M, and if so, the local optimal solution is discarded, and a new solution is generated according to the following formula: x ij = x mini + a(x maxj - x minj ) where: a is a random number between 0 and 1 ; x mini , x maxj , x minj are the minimum and maximum values of the corresponding dimension solution; then repeat the above steps; In the fifth step, it is checked whether the maximum number of iterations 300 has been reached, and if so, the optimal solution can be output. Finally, the optimal output value of the thermal storage combined frequency modulation system is obtained, and the trend graph of the state of charge of the energy storage battery is further obtained.

6. The method of claim 1, wherein, In the step S5, the optimal output value of the thermal storage obtained by the optimization calculation is applied to the variable load process of the thermal storage combined frequency modulation, so that the total economic benefit of the thermal storage combined frequency modulation system tends to be theoretically optimal when frequency modulation is performed; when applying, the load instruction prediction value issued in advance by the power grid side needs to be compared with the actual load instruction value, i.e., when the deviation between the day-ahead prediction load instruction and the actual load instruction value is less than a certain allowable value L, the previously calculated optimal output value of the thermal storage is applied to the variable load process with smaller load instruction deviation, which is equivalent to performing optimization calculation in advance; when the deviation between the day-ahead prediction load instruction and the actual load instruction is greater than the allowable value L, the deviation between the previously optimized calculation result and the actual optimal output value of the thermal storage exceeds the allowable range, and at this time, the frequency modulation instruction response is performed according to the current thermal storage output matching mode of the corresponding power plant, i.e., at this time, the previously calculated optimal output value of the thermal storage cannot be used.

7. The method of claim 1, wherein the method further comprises: In the step S6, the state of charge of the energy storage battery is adjusted in advance, provided that the thermal power unit and the energy storage battery have responded to the previous frequency modulation instruction, the frequency modulation performance evaluation index of the power grid side has been completely evaluated, and a new frequency modulation instruction has not appeared; at this time, the thermal power unit will make small power fluctuations around the required power due to inertia, and the fluctuating power is used to pre-charge and discharge the energy storage, i.e., to adjust the state of charge.

8. The method of claim 7, wherein the optimization is performed by a genetic algorithm. The pre-charge and discharge is based on the trend graph of the state of charge of the energy storage battery obtained by the previous optimization calculation, and the direction and size of the change in the state of charge at the next time point in the trend graph are used to adjust the current state of charge, with a time point every 5 minutes, and the adjustment target is to make the state of charge of the energy storage far away from the upper and lower limits of 0.2 or 0.8, and to improve the availability of the energy storage.

Images