Wind power output control method and system connected to transformer substation of thermal power plant

Abstract

The invention discloses a wind power output control method and system accessed to a transformer substation of a thermal power plant, and relates to the technical field of wind power output control, and the method comprises the steps: obtaining multi-source data, including ultra-short-term wind power prediction data, real-time operation data and adjustable potential data; performing cost analysis on the real-time operation data to obtain deep adjustment cost of the target thermal power generating unit; constructing a hierarchical cooperative control architecture which comprises an upper-layer optimization model and a lower-layer optimization model; performing collaborative analysis on the ultra-short-term wind power prediction data and the adjustable potential data by taking the minimum wind curtailment amount as a target to obtain an upper-layer scheduling decision; based on the upper-layer scheduling decision, total cost is obtained according to the deep scheduling cost, and a lower-layer scheduling decision is obtained by taking the minimum total cost as a target and comprises a start-stop output scheme; and controlling the target thermal power generating unit according to the start-stop output scheme. The technical problem that in the prior art, the wind curtailment amount is large when wind power is connected into a thermal power plant substation is solved, and the technical effect of reducing the wind curtailment amount is achieved.

Description

Technical Field

[0001] This invention relates to the field of wind power output control technology, specifically to a wind power output control method and system connected to a thermal power plant substation. Background Technology

[0002] In the application scenario of wind power being connected to thermal power plant substations, wind power output is fluctuating and random, while the output regulation of thermal power units is usually limited by peak-shaving capacity, start-up and shutdown constraints and operating conditions. The adjustable resources of the grid-side load have not been coordinated and scheduled in a unified manner with wind power forecast information and thermal power operating status. As a result, when wind power generation is high and the system absorption capacity is insufficient, wind power output is difficult to be effectively absorbed in a timely manner. The system balance can only be maintained by limiting wind power generation and curtailing wind power, thus causing a large amount of wind curtailment. Summary of the Invention

[0003] This application provides a method and system for controlling wind power output when connected to a thermal power plant substation, which is used to address the technical problem of large wind curtailment when wind power is connected to a thermal power plant substation in the prior art.

[0004] In view of the above problems, this application provides a method and system for controlling wind power output connected to a thermal power plant substation.

[0005] The first aspect of this application provides a method for controlling wind power output connected to a thermal power plant substation, the method comprising: The process involves acquiring multi-source data, including ultra-short-term wind power forecast data for the target wind farm, real-time operating data of the target thermal power unit, and adjustable potential data of the target grid-side load. A deep-schedule cost contingency plan is introduced to analyze the cost of the real-time operating data, yielding the deep-schedule cost of the target thermal power unit. A hierarchical collaborative control architecture is constructed, comprising an upper-level optimization model and a lower-level optimization model. The upper-level optimization model, aiming to minimize wind curtailment, performs collaborative analysis on the ultra-short-term wind power forecast data and the adjustable potential data to obtain an upper-level scheduling decision. Based on the upper-level scheduling decision and the deep-schedule cost, the lower-level optimization model obtains a total cost and aims to minimize the total cost, resulting in a lower-level scheduling decision, which includes the start-stop output scheme for the target thermal power unit. The target thermal power unit is then controlled according to the start-stop output scheme.

[0006] A second aspect of this application provides a wind power output control system connected to a thermal power plant substation, the system comprising: The system includes a data acquisition module for acquiring multi-source data, including ultra-short-term wind power forecast data of the target wind farm, real-time operation data of the target thermal power unit, and adjustable potential data of the target grid-side load; a cost analysis module for introducing a deep-schedule cost plan to perform cost analysis on the real-time operation data to obtain the deep-schedule cost of the target thermal power unit; an architecture construction module for constructing a hierarchical collaborative control architecture, including an upper-level optimization model and a lower-level optimization model; a collaborative analysis module for the upper-level optimization model to perform collaborative analysis on the ultra-short-term wind power forecast data and the adjustable potential data with the goal of minimizing wind curtailment, to obtain an upper-level scheduling decision; a lower-level scheduling decision determination module for the lower-level optimization model to obtain a total cost based on the upper-level scheduling decision and the deep-schedule cost, with the goal of minimizing the total cost, to obtain a lower-level scheduling decision, including the start-stop output scheme of the target thermal power unit; and a control module for controlling the target thermal power unit according to the start-stop output scheme.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages: This application acquires multi-source data, including ultra-short-term wind power forecast data of the target wind farm, real-time operation data of the target thermal power unit, and adjustable potential data of the target grid-side load. A deep-schedule cost contingency plan is introduced to analyze the cost of the real-time operation data, yielding the deep-schedule cost of the target thermal power unit. A hierarchical collaborative control architecture is constructed, comprising an upper-level optimization model and a lower-level optimization model. The upper-level optimization model, with the goal of minimizing wind curtailment, performs collaborative analysis on the ultra-short-term wind power forecast data and the adjustable potential data to obtain an upper-level scheduling decision. The lower-level optimization model, based on the upper-level scheduling decision and the deep-schedule cost, obtains the total cost and, with the goal of minimizing the total cost, obtains a lower-level scheduling decision, including the start-stop output scheme of the target thermal power unit. The target thermal power unit is controlled according to the start-stop output scheme. This invention addresses the technical problem of large wind curtailment when wind power is connected to thermal power plant substations in the prior art. By constructing a hierarchical collaborative control architecture including an upper-level optimization model and a lower-level optimization model, it coordinates the scheduling of ultra-short-term wind power forecast data, real-time operating data of thermal power units, and grid-side load adjustment potential data, thereby achieving the technical effect of reducing wind curtailment. Attached Figure Description

[0008] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0009] Figure 1 A schematic flowchart of a wind power output control method for connecting to a thermal power plant substation is provided in an embodiment of this application. Figure 2 This is a schematic diagram of a wind power output control system connected to a thermal power plant substation, provided as an embodiment of this application.

[0010] Figure labeling: Data acquisition module 11, cost analysis module 12, architecture construction module 13, collaborative analysis module 14, lower-level scheduling decision determination module 15, control module 16. Detailed Implementation

[0011] This application provides a method and system for controlling wind power output connected to a thermal power plant substation. Addressing the technical problem of significant wind curtailment when wind power is connected to a thermal power plant substation in existing technologies, this method constructs a hierarchical collaborative control architecture including an upper-level optimization model and a lower-level optimization model. This architecture collaboratively schedules ultra-short-term wind power forecast data, real-time operating data of thermal power units, and grid-side load adjustment potential data, thereby achieving the technical effect of reducing wind curtailment.

[0012] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0013] It should be noted that any variation of the terms "comprising" and "having" is intended to cover non-exclusive inclusion, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to such processes, methods, products, or devices.

[0014] Example 1, as Figure 1 As shown, this application provides a method for controlling wind power output connected to a thermal power plant substation, the method comprising: Step S100: Obtain multi-source data, wherein the multi-source data includes ultra-short-term wind power prediction data of the target wind farm, real-time operation data of the target thermal power unit, and adjustable potential data of the target grid-side load.

[0015] In the embodiments of this application, the multi-source data includes ultra-short-term wind power forecast data of the target wind farm, real-time operation data of the target thermal power unit, and adjustable potential data of the target grid-side load.

[0016] The ultra-short-term wind power prediction data for the target wind farm is obtained through a wind power prediction model. Specifically, wind speed, wind direction, ambient temperature, and actual wind power data of the target wind farm during its historical operation are first collected and arranged at fixed time intervals. Then, the wind speed, wind direction, ambient temperature, and actual wind power data of the previous time period are used as input data, and the actual wind power of the next time period is used as output data, forming multiple sets of training samples. The training samples are input into the wind power prediction model for training, and the model parameters are adjusted according to the deviation between the prediction results and the actual wind power until the deviation meets the preset requirements, thus completing the training of the wind power prediction model. In actual use, the wind speed, wind direction, ambient temperature, and current actual wind power collected at the target wind farm at the current moment are input into the trained wind power prediction model, which outputs the predicted power values ​​for each moment in the preset future time period. These predicted power values ​​are used as the ultra-short-term wind power prediction data for the target wind farm. The ultra-short-term wind power prediction data refers to the predicted output data for the next scheduling cycle or several consecutive short-term scheduling cycles.

[0017] The real-time operating data of the target thermal power unit is generated by reading the current operating status of the target thermal power unit. Specifically, the unit load, start-up and shutdown status, minimum stable combustion load, load increase / decrease capacity, and grid connection status at the current moment are read from the unit monitoring device corresponding to the target thermal power unit, and the read data are recorded at the same time to obtain the real-time operating data of the target thermal power unit; where real-time operating data refers to data that can characterize the current operating condition of the target thermal power unit.

[0018] The adjustable potential data of the target grid-side load is generated by calculating the adjustment range of each load unit. Specifically, the current load power, maximum allowable load power, and minimum allowable load power of each load unit are read; the maximum allowable load power is subtracted from the current load power to obtain the adjustable power of the corresponding load unit; the current load power is subtracted from the minimum allowable load power to obtain the adjustable power of the corresponding load unit; then the adjustable power and adjustable power of each load unit are summarized to obtain the adjustable potential data of the target grid-side load. Among them, the adjustable potential data refers to the capacity data of the target grid-side load that can adjust power within the current time period.

[0019] After obtaining the ultra-short-term wind power forecast data of the target wind farm, the real-time operation data of the target thermal power unit, and the adjustable potential data of the target grid-side load, the three types of data are organized according to the same scheduling time to form a data group corresponding to that scheduling time, thereby obtaining multi-source data.

[0020] Step S200: Introduce a deep adjustment cost plan to perform cost analysis on the real-time operating data to obtain the deep adjustment cost of the target thermal power unit.

[0021] In this embodiment of the application, when the deep adjustment cost plan is introduced to perform cost analysis on real-time operating data, the real-time operating data is subjected to variation weighted calculation to obtain the real-time operating coefficient reflecting the current operating status of the target thermal power unit; the initial stable combustion load of the target thermal power unit is corrected with the real-time operating coefficient as the weight to obtain the minimum stable combustion load; then, the minimum stable combustion load is used as the peak-shaving boundary to divide the operating range of the target thermal power unit into peak-shaving sections to obtain the target division results; finally, the deep peak-shaving cost analysis is performed on the target division results according to the deep adjustment cost plan to obtain the deep adjustment cost of the target thermal power unit.

[0022] Furthermore, in the method provided in the application embodiment, a deep adjustment cost contingency plan is introduced to perform cost analysis on the real-time operating data to obtain the deep adjustment cost of the target thermal power unit, which also includes: The real-time operating data is subjected to variation weighting calculation to obtain the real-time operating coefficient; the initial stable combustion load of the target thermal power unit is adjusted with the real-time operating coefficient as the weight to obtain the minimum stable combustion load; the target thermal power unit is divided into peak shaving categories based on the minimum stable combustion load to obtain the target category results; the target category results are analyzed for deep peak shaving cost according to the deep shaving cost plan to obtain the deep shaving cost.

[0023] In this embodiment, when performing variation-weighted calculation on real-time operating data, the unit load, load increase / decrease capacity, and combustion state parameters of the target thermal power unit at the current moment are first extracted from the real-time operating data, and the corresponding unit load benchmark value, load increase / decrease capacity benchmark value, and combustion state parameter benchmark value are read respectively. Then, the same difference processing method is used for calculation, that is, the unit load difference, load increase / decrease capacity difference, and combustion state parameter difference are obtained by subtracting the current value from the benchmark value. Subsequently, the unit load difference is divided by the unit load benchmark value to obtain the unit load variation value, the load increase / decrease capacity difference is divided by the load increase / decrease capacity benchmark value to obtain the load increase / decrease capacity variation value, and the combustion state parameter difference is divided by the combustion state parameter benchmark value to obtain the combustion state parameter variation value. Then, the unit load variation value is multiplied by a preset unit load weight, the load increase / decrease capacity variation value is multiplied by a preset load increase / decrease capacity weight, and the combustion state parameter variation value is multiplied by a preset combustion state parameter weight. The three product results are added together, and then a base coefficient of 1 is added to the sum to obtain the real-time operating coefficient.

[0024] Next, when adjusting the initial stable combustion load of the target thermal power unit with the real-time operating coefficient as the weight, the initial stable combustion load of the target thermal power unit is first read, and then the real-time operating coefficient obtained in the previous step is read; then the initial stable combustion load is multiplied by the real-time operating coefficient to obtain the corrected load value; then the corrected load value is determined as the minimum stable combustion load.

[0025] Then, the target thermal power units are divided into peak shaving stages based on the minimum stable combustion load. First, the current actual operating load and oil injection status of the target thermal power units are read. Then, the actual operating load is compared with the initial stable combustion load and the minimum stable combustion load. When the actual operating load is less than the initial stable combustion load and greater than or equal to the minimum stable combustion load, and the oil injection status is not oil injection, the current operating status is divided into the deep peak shaving stage without oil injection. When the actual operating load is less than the minimum stable combustion load, and the oil injection status is oil injection, the current operating status is divided into the deep peak shaving stage with oil injection. After completing the division according to the above comparison and judgment steps, the target division result is obtained, and the target division result includes the deep peak shaving stage without oil injection and the deep peak shaving stage with oil injection.

[0026] Finally, a deep peak-shaving cost analysis is conducted on the target allocation results based on the deep peak-shaving cost contingency plan. Specifically, the deep peak-shaving cost is determined according to the peak-shaving stage corresponding to the target allocation results. When the target allocation results include a deep peak-shaving stage without oil injection, the rotor damage cost of the target thermal power unit is calculated according to the deep peak-shaving cost contingency plan, and the rotor damage cost is used as the deep peak-shaving cost. When the target allocation results include a deep peak-shaving stage with oil injection, the fuel cost of the target thermal power unit is calculated according to the deep peak-shaving cost contingency plan, and the sum of the rotor damage cost and the fuel cost is used as the deep peak-shaving cost.

[0027] Furthermore, the method provided in the application embodiments also includes: The target segmentation result includes a deep peak shaving stage without oil injection. The deep peak shaving cost analysis is performed on the target segmentation result according to the deep peak shaving cost plan to obtain the deep peak shaving cost, including: calculating the rotor damage cost of the target thermal power unit in the deep peak shaving stage without oil injection according to the deep peak shaving cost plan, and using the rotor damage cost as the deep peak shaving cost.

[0028] In this embodiment of the application, when the target division result includes a deep peak shaving phase without oil injection, when performing a deep peak shaving cost analysis on the target division result according to the deep peak shaving cost plan, the phase operation data of the target thermal power unit in the deep peak shaving phase without oil injection is extracted. The phase operation data includes the actual operating load, load change amplitude, load change rate, and continuous operating duration. At the same time, the loss parameter table corresponding to the deep peak shaving phase without oil injection in the deep peak shaving cost plan is retrieved. The loss parameter table includes at least the basic loss rate corresponding to each load interval, the amplitude correction coefficient corresponding to each load change amplitude interval, and the rate correction coefficient corresponding to each load change rate interval. The actual operating load is matched with the load range in the loss parameter table to determine the basic loss rate; the load change amplitude is matched with the amplitude range in the loss parameter table to determine the amplitude correction coefficient; the load change rate is matched with the rate range in the loss parameter table to determine the rate correction coefficient; then the basic loss rate is multiplied by the amplitude correction coefficient, and then multiplied by the rate correction coefficient to obtain the loss value per unit time; finally, the loss value per unit time is multiplied by the continuous operating time to obtain the rotor loss of the target thermal power unit during the peak shaving phase without oil injection.

[0029] After obtaining the rotor damage amount, the pre-stored unit damage cost is read from the deep adjustment cost plan, and then the rotor damage amount is multiplied by the unit damage cost to obtain the rotor damage cost of the target thermal power unit; after completing the rotor damage cost calculation, the rotor damage cost is directly output as the deep adjustment cost.

[0030] Furthermore, the method provided in the application embodiments also includes: The target segmentation result includes the oil injection depth peak shaving stage. According to the deep peak shaving cost plan, the target segmentation result is analyzed for deep peak shaving cost to obtain the deep peak shaving cost, including: according to the deep peak shaving cost plan, in the oil injection depth peak shaving stage, the fuel cost of the target thermal power unit is calculated; the sum of the rotor damage cost and the fuel cost is taken as the deep peak shaving cost.

[0031] In this embodiment of the application, when the target division result includes the deep peak shaving stage of fuel injection, when performing deep peak shaving cost analysis on the target division result according to the deep peak shaving cost plan, the deep peak shaving stage of fuel injection is first identified from the target division result, and the fuel injection operation data of the target thermal power unit in the deep peak shaving stage of fuel injection is extracted; the fuel injection operation data includes fuel injection flow rate, fuel injection duration and fuel price per unit. The deep peak shaving stage of fuel injection is the operation stage in which the target thermal power unit needs to inject fuel to maintain combustion during the deep peak shaving process. Subsequently, the fuel calculation rules corresponding to the current fuel injection method are read according to the deep adjustment cost plan, and the fuel consumption is calculated according to the fuel injection flow rate and fuel injection duration. When the fuel injection flow rate remains constant during the deep fuel injection peak shaving phase, the fuel injection flow rate and fuel injection duration are multiplied to obtain the fuel consumption. When the fuel injection flow rate changes during the deep fuel injection peak shaving phase, the fuel injection flow rate is recorded separately for each sampling period, and the fuel injection flow rate of each sampling period is multiplied by the corresponding duration and then summed to obtain the fuel consumption. After obtaining the fuel consumption, the fuel consumption is multiplied by the fuel unit price to obtain the fuel cost of the target thermal power unit. The fuel cost is the cost value incurred by the target thermal power unit for fuel injection combustion assistance during the deep fuel injection peak shaving phase.

[0032] After obtaining the fuel cost, the rotor damage cost of the target thermal power unit during the deep peak shaving process is read, and the rotor damage cost is summed with the fuel cost to obtain the deep peak shaving cost corresponding to the deep peak shaving stage. Among them, the rotor damage cost is the damage cost of the target thermal power unit's rotor during the deep peak shaving process, and the deep peak shaving cost is the total cost of implementing deep peak shaving during the deep peak shaving stage of the target thermal power unit.

[0033] Step S300: Construct a hierarchical collaborative control architecture, wherein the hierarchical collaborative control architecture includes an upper-level optimization model and a lower-level optimization model.

[0034] In this embodiment, when constructing the hierarchical collaborative control architecture, the ultra-short-term wind power forecast data of the target wind farm, the real-time operation data of the target thermal power unit, and the adjustable potential data of the target grid load are first organized according to the same scheduling cycle, and the three types of data corresponding to each scheduling moment are combined into a set of scheduling data. Then, according to the decision-making order in the scheduling process, the overall scheduling process is divided into upper-level decision-making and lower-level decision-making, and a hierarchical collaborative control architecture is established accordingly. The hierarchical collaborative control architecture includes an upper-level optimization model and a lower-level optimization model. The upper-level optimization model uses ultra-short-term wind power forecast data and adjustable potential data as input data, and sets the minimum wind curtailment as the optimization objective. Then, the power balance relationship, the adjustment boundary of the target grid load, and the scheduling period constraints are written into the upper-level optimization model, and the upper-level scheduling decision is obtained after calculation. The upper-level scheduling decision includes the adjustment power undertaken by the target grid load in each scheduling period and the remaining adjustment demand to be undertaken by the target thermal power unit.

[0035] After obtaining the upper-level scheduling decision, the upper-level scheduling decision and the real-time operating data of the target thermal power unit are used together as the input data of the lower-level optimization model. The coal consumption cost, deep adjustment cost, and start-up and shutdown cost of the target thermal power unit are summarized to form the total cost. The minimum total cost is then set as the optimization objective of the lower-level optimization model. At the same time, the start-up and shutdown state constraints, output upper and lower limit constraints, and scheduling power constraints of the target thermal power unit are written into the lower-level optimization model. After calculation, the lower-level scheduling decision is obtained, which includes the start-up and shutdown output scheme of the target thermal power unit. Through the above process, the upper-level optimization model and the lower-level optimization model are established, and the data transmission relationship from the upper-level scheduling decision output to the lower-level optimization model input is established. Thus, the construction of the hierarchical collaborative control architecture is completed.

[0036] Step S400: The upper-level optimization model takes minimizing wind curtailment as its objective, and performs collaborative analysis on the ultra-short-term wind power prediction data and the adjustable potential data to obtain upper-level scheduling decisions.

[0037] In this embodiment, the upper-level optimization model aims to minimize wind curtailment. When collaboratively analyzing ultra-short-term wind power forecast data and adjustable potential data, a hierarchical dispatch strategy is introduced. The upper-level dispatch decision is determined according to the order of priority adjustment of the target grid-side load and subsequent peak shaving of the target thermal power units. When the peak shaving capacity is insufficient due to large wind power generation, the upward adjustment power of the target grid-side load is prioritized in the zero-cost adjustment range to form a first load power adjustment plan, which is used as the upper-level dispatch decision. When the zero-cost adjustment range is exhausted, the real-time adjustable capacity of the target grid-side load is dynamically calculated based on the power-temperature coupling constraint of thermal inertia to form a second load power adjustment plan, which is used as the upper-level dispatch decision. When both the zero-cost adjustment range and the real-time adjustable capacity are exhausted, deep peak shaving is performed on the target thermal power units to form a total output plan for the thermal power units, which is used as the upper-level dispatch decision.

[0038] Furthermore, in the method provided in the application embodiments, the upper-level optimization model, with the goal of minimizing wind curtailment, performs collaborative analysis on the ultra-short-term wind power forecast data and the adjustable potential data to obtain upper-level scheduling decisions, and further includes: A tiered allocation strategy is introduced, aiming to minimize wind curtailment. This involves collaborative analysis of the ultra-short-term wind power forecast data and the adjustable potential data to determine the upper-level dispatch decision. The tiered allocation strategy includes: when a surge in wind power generation leads to insufficient peak-shaving capacity, configuring the increased power of the target grid-side load within a zero-cost adjustment range to form a first load power adjustment plan, which is then used as the upper-level dispatch decision; when the zero-cost adjustment range is exhausted, dynamically calculating the real-time adjustable capacity of the target grid-side load based on the power-temperature coupling constraint of thermal inertia to form a second load power adjustment plan, which is then used as the upper-level dispatch decision; and when both the zero-cost adjustment range and the real-time adjustable capacity are exhausted, performing deep peak shaving on the target thermal power units to form a total output plan for the thermal power units, which is then used as the upper-level dispatch decision.

[0039] In this embodiment, a hierarchical allocation strategy is introduced. With the goal of minimizing wind curtailment, the ultra-short-term wind power forecast data and adjustable potential data are analyzed collaboratively to determine the upper-level scheduling decision. First, the ultra-short-term wind power forecast data and adjustable potential data are aligned according to a unified scheduling cycle. Within each scheduling period, the predicted wind power value of the target wind farm and the adjustable power and the upper limit of the zero-cost adjustment range corresponding to each load unit of the target grid-side load are extracted. Subsequently, within each scheduling period, the allocated wind power acceptance power and the remaining peak-shaving capacity of the target thermal power unit are read. The peak-shaving gap value is calculated by subtracting the allocated wind power acceptance power from the predicted wind power value and then subtracting the remaining peak-shaving capacity of the target thermal power unit. When the peak-shaving gap value is greater than zero, it is determined that the peak-shaving capacity is insufficient due to the large wind power generation in the current scheduling period, and the peak-shaving gap value is determined as the power to be allocated in the current scheduling period. Then, the zero-cost adjustment range of each load unit is allocated according to a preset order. First, the adjustable power and the upper limit of the zero-cost adjustment range of the current load unit are read. When the adjustable power is less than or equal to the upper limit of the zero-cost adjustment range, the adjustable power is determined as the zero-cost allocable power of that load unit. When the adjustable power is greater than the upper limit of the zero-cost adjustment range, the upper limit of the zero-cost adjustment range is determined as the zero-cost allocable power of that load unit. Next, the power to be allocated is compared with the zero-cost allocable power of the current load unit. When the power to be allocated is less than or equal to the zero-cost allocable power, all the power to be allocated is allocated to the current load unit. The increased power of the current load unit is recorded as the power to be allocated, and the allocation for the current scheduling period ends. When the power to be allocated is greater than the zero-cost allocable power, all the zero-cost allocable power is allocated to the current load unit, and the power to be allocated is subtracted from the zero-cost allocable power to obtain the new power to be allocated. The same process is then repeated for the next load unit until the power to be allocated is reduced to zero or the zero-cost adjustment range of all load units has been allocated. After the allocation for each scheduling period is completed, the increased power obtained by each load unit in each scheduling period is summarized in chronological order to form the first load power adjustment plan, and the first load power adjustment plan is used as the upper-level scheduling decision.

[0040] Once the zero-cost adjustment range is exhausted, the real-time adjustable capacity of the target grid-side load is dynamically calculated based on the power-temperature coupling constraint of thermal inertia. In this process, the maximum temperature adjustment is determined based on the actual temperature of the target grid-side load and the preset safe temperature. Then, a predetermined factor is retrieved, and combined with the maximum temperature adjustment, the maximum allowable adjustment power is determined. Subsequently, using the maximum allowable adjustment power as the adjustment constraint, power is allocated to the target grid-side load to form a second load power adjustment plan, which is then used as the upper-level scheduling decision.

[0041] When both the zero-cost adjustment range and the real-time adjustable capacity are exhausted, and deep peak shaving is performed on the target thermal power unit, the remaining power that has not been absorbed after load adjustment by the target grid side within the current dispatch period is first read, and this remaining power is determined as the peak-shaving power of the target thermal power unit within the current dispatch period. Then, the current actual output, minimum stable combustion load, and adjustable output range of the target thermal power unit are read, and the target adjustment output is calculated by subtracting the peak-shaving power from the current actual output. When the target adjustment output is greater than or equal to the minimum stable combustion load, The output of the target thermal power unit is adjusted to the target adjusted output, and the target adjusted output is recorded in the output value of the thermal power unit corresponding to the current scheduling period. When the target adjusted output is less than the minimum stable combustion load, the output of the target thermal power unit is adjusted to the minimum stable combustion load, and the minimum stable combustion load is recorded in the output value of the thermal power unit corresponding to the current scheduling period. Then, the same processing is performed on each scheduling period, and the output values ​​of the thermal power units corresponding to each scheduling period are summarized in time sequence to form the total output plan of the thermal power units. The total output plan of the thermal power units is then used as the upper-level scheduling decision.

[0042] Furthermore, in the method provided in the application embodiment, after the zero-cost adjustment range is exhausted, the real-time adjustable capacity of the target grid-side load is dynamically calculated based on the power-temperature coupling constraint of thermal inertia to form a second load power adjustment plan, which further includes: The maximum temperature adjustment amount is obtained based on the actual temperature of the target grid-side load and the preset safe temperature; a predetermined factor is retrieved and combined with the maximum temperature adjustment amount to determine the maximum allowable adjustment power; the second load power adjustment plan is formed with the maximum allowable adjustment power as a constraint.

[0043] In this embodiment, when obtaining the maximum temperature adjustment amount based on the actual temperature of the target grid-side load and the preset safe temperature, the actual temperature of the target grid-side load during the current scheduling period is first read, and the preset safe temperature corresponding to the target grid-side load is also read. The actual temperature is the current temperature value detected by the temperature monitoring device, and the preset safe temperature is a pre-set temperature allowable boundary value. Then, the actual temperature is compared with the preset safe temperature. When the preset safe temperature is greater than the actual temperature, the maximum temperature adjustment amount is calculated by subtracting the actual temperature from the preset safe temperature; when the actual temperature is greater than the preset safe temperature, the maximum temperature adjustment amount is calculated by subtracting the preset safe temperature from the actual temperature. After the calculation is completed, the maximum temperature adjustment amount corresponding to the target grid-side load during the current scheduling period is obtained.

[0044] When retrieving a predetermined factor and combining it with the maximum temperature adjustment to determine the maximum allowable adjustment power, the predetermined factor corresponding to the target grid load is first read. This predetermined factor represents the power change corresponding to a unit temperature change. Then, the maximum temperature adjustment is multiplied by the predetermined factor to obtain the maximum allowable adjustment power corresponding to the target grid load during the current dispatch period. That is, when the maximum temperature adjustment increases, the calculated maximum allowable adjustment power increases accordingly; when the maximum temperature adjustment decreases, the calculated maximum allowable adjustment power decreases accordingly. Through the above calculation, the conversion from temperature change to power adjustment is completed, yielding the maximum allowable adjustment power.

[0045] Finally, when forming the second load power regulation plan with the maximum allowable adjustment power as a constraint, the remaining power to be allocated after the zero-cost regulation interval is exhausted is read first, and this power to be allocated is taken as the regulation power that needs to be allocated in the current scheduling period. Then, the power to be allocated is compared with the maximum allowable adjustment power. When the power to be allocated is less than or equal to the maximum allowable adjustment power, the power to be allocated is determined as the regulation power of the target grid-side load in the current scheduling period. When the power to be allocated is greater than the maximum allowable adjustment power, the maximum allowable adjustment power is determined as the regulation power of the target grid-side load in the current scheduling period. After that, the regulation power determined in each scheduling period is summarized in chronological order to form the second load power regulation plan.

[0046] Step S500: The lower-level optimization model obtains the total cost based on the upper-level scheduling decision and the deep adjustment cost, and obtains the lower-level scheduling decision with the goal of minimizing the total cost. The lower-level scheduling decision includes the start-up and shutdown output scheme of the target thermal power unit.

[0047] Furthermore, the method provided in the application embodiments also includes: The total cost includes at least the coal consumption cost of the target thermal power unit, the deep adjustment cost, and the start-up and shutdown cost.

[0048] In this embodiment, the lower-level optimization model is based on the upper-level scheduling decision. It obtains the total cost based on the deep adjustment cost and aims to minimize the total cost. When obtaining the lower-level scheduling decision, it first reads the upper-level scheduling decision and the real-time operation data of the target thermal power unit corresponding to each scheduling period according to the unified scheduling cycle. Then, it determines the adjustment demand allocated to the target thermal power unit in the upper-level scheduling decision within each scheduling period as the target output demand within that scheduling period. At the same time, it reads the current start-stop status, current output, minimum stable combustion load, initial stable combustion load, output upper limit, and allowable output variation range of the target thermal power unit within that scheduling period. The above data is written into the calculation table in the order of scheduling periods to form the time-series input data of the lower-level optimization model.

[0049] After generating the time-series input data, two start / stop states are set for each scheduling period: power-on and power-off. In the power-on state, the target output demand is written as a candidate output value for the current scheduling period. When the target output demand is less than the minimum stable fuel load, the minimum stable fuel load is written as the candidate output value for the current scheduling period. When the target output demand is greater than the output limit, the output limit is written as the candidate output value for the current scheduling period. When the target output demand is between the minimum stable fuel load and the output limit, the target output demand is directly written as the candidate output value for the current scheduling period. In the power-off state, the candidate output value for the current scheduling period is written as zero. After completing the writing of the start / stop states and candidate output values ​​for each scheduling period, multiple candidate start / stop output schemes are obtained.

[0050] After obtaining multiple candidate start-up and shutdown power output schemes, the validity of each candidate scheme is verified by comparing the difference in power output values ​​between adjacent scheduling periods and comparing this difference with the allowable range of power output variation. If the difference in power output values ​​between adjacent scheduling periods is greater than the allowable range of power output variation, the corresponding candidate start-up and shutdown power output scheme is deleted. If the power output value of a candidate in the start-up state is less than the minimum stable fuel load or greater than the upper limit of power output, the corresponding candidate start-up and shutdown power output scheme is deleted. If the candidate start-up and shutdown power output scheme is inconsistent with the target power output demand corresponding to the upper-level scheduling decision and cannot meet the adjustment demand by adjusting to the minimum stable fuel load or the upper limit of power output, the corresponding candidate start-up and shutdown power output scheme is deleted. After screening, the candidate start-up and shutdown power output schemes that meet the start-up and shutdown state constraints, upper and lower limit of power output constraints, and scheduling power constraints are retained as the objects of total cost calculation.

[0051] After retaining the candidate start-up and shutdown output schemes that meet the constraints, the coal consumption cost is calculated. This involves reading the candidate output values ​​for each scheduling period one by one, and looking up the unit coal consumption value corresponding to the candidate output value according to the pre-stored coal consumption correspondence table. Then, the unit coal consumption value is multiplied by the output duration of the current scheduling period to obtain the coal consumption amount for the current scheduling period. Subsequently, the coal consumption amount is multiplied by the coal price to obtain the coal consumption cost for the current scheduling period. The coal consumption costs for each scheduling period are accumulated to obtain the coal consumption cost of the target thermal power unit corresponding to the current candidate start-up and shutdown output scheme.

[0052] After calculating the coal consumption cost, the deep adjustment cost is calculated. Candidate output values ​​for each scheduling period are read one by one, and these values ​​are compared with the initial stable combustion load and the minimum stable combustion load. When the candidate output value is less than the initial stable combustion load but greater than or equal to the minimum stable combustion load, the current scheduling period is determined to have entered the deep peak shaving phase without oil injection, and the corresponding rotor damage cost is read as the deep adjustment cost for the current scheduling period. When the candidate output value is less than the minimum stable combustion load, the current scheduling period is determined to have entered the deep peak shaving phase with oil injection, and the corresponding rotor damage cost and fuel cost are read. The rotor damage cost and fuel cost are added together as the deep adjustment cost for the current scheduling period. When the candidate output value is greater than or equal to the initial stable combustion load, the deep adjustment cost for the current scheduling period is recorded as zero. The deep adjustment costs for each scheduling period are accumulated to obtain the deep adjustment cost corresponding to the current candidate start-stop output scheme.

[0053] After calculating the deep adjustment cost, the start-up and shutdown cost is calculated. The start-up and shutdown statuses in adjacent scheduling periods are compared one by one. When the previous scheduling period is in a shutdown state and the current scheduling period is in an operating state, the pre-stored operating cost is read as the start-up and shutdown cost for the current scheduling period. When the previous scheduling period is in an operating state and the current scheduling period is in a shutdown state, the pre-stored shutdown cost is read as the start-up and shutdown cost for the current scheduling period. When the start-up and shutdown statuses of adjacent scheduling periods do not change, the start-up and shutdown cost for the current scheduling period is recorded as zero. The start-up and shutdown costs of each scheduling period are accumulated to obtain the start-up and shutdown cost corresponding to the current candidate start-up and shutdown output scheme.

[0054] After obtaining the coal consumption cost, deep adjustment cost, and start-up / shutdown cost respectively, for each candidate start-up / shutdown output scheme, the total cost is calculated by adding the coal consumption cost, deep adjustment cost, and start-up / shutdown cost. Then, the total costs corresponding to all candidate start-up / shutdown output schemes are compared one by one, and the candidate start-up / shutdown output scheme with the smallest total cost is selected as the output result of the lower-level optimization model, and this output result is determined as the lower-level scheduling decision. The lower-level scheduling decision includes the start-up / shutdown output scheme of the target thermal power unit, and the start-up / shutdown output scheme includes the start-up / shutdown status and corresponding output value in each scheduling period.

[0055] Step S600: Control the target thermal power unit according to the start-stop power output scheme.

[0056] In this embodiment, when controlling the target thermal power unit according to the start-stop output scheme, the start-stop status and output value corresponding to each scheduling period in the start-stop output scheme are first read. Then, the start-stop status is sent to the start-stop control terminal of the target thermal power unit, and the output value is sent to the load control terminal of the target thermal power unit. When the start-stop status is the start-up status, the target thermal power unit is controlled to start operation, and the unit output is adjusted to the corresponding output value. When the start-stop status is the shutdown status, the target thermal power unit is controlled to stop operation. Then, the corresponding start-stop control and output adjustment are executed sequentially according to the timing of each scheduling period, thus completing the control of the target thermal power unit according to the start-stop output scheme.

[0057] Furthermore, in the method provided in the application embodiment, after controlling the target thermal power unit according to the start-stop power output scheme, it further includes: Acquire control and monitoring information, including the subsynchronous torsional vibration frequency of the target thermal power unit and the shaft torsional vibration frequency of the target wind farm; issue a risk warning when the sum of the subsynchronous torsional vibration frequency and the shaft torsional vibration frequency reaches a predetermined grid fundamental frequency; retrieve the risk emergency plan based on the risk warning to carry out emergency handling of the target wind farm; wherein, the risk emergency plan refers to adjusting the current inner loop gain coefficient of the rotor-side frequency converter of the target wind farm based on a predetermined step size.

[0058] In this embodiment, when acquiring control monitoring information, a vibration monitoring device or an electrical quantity sampling device is first arranged at the shaft position of the target thermal power unit, and a vibration monitoring device or an electrical quantity sampling device is arranged at the wind turbine main shaft, generator side, or rotor side inverter position of the target wind farm. Then, the monitoring signals of the target thermal power unit and the target wind farm are synchronously sampled according to a unified sampling period to obtain the corresponding time-domain sampling data. Then, the time-domain sampling data of the target thermal power unit and the time-domain sampling data of the target wind farm are respectively subjected to discrete Fourier transform to obtain their respective spectrum data. Then, the frequency point with the largest spectrum amplitude is read within a preset frequency range, and the frequency point corresponding to the target thermal power unit is determined as the subsynchronous torsional vibration frequency of the target thermal power unit, and the frequency point corresponding to the target wind farm is determined as the shaft torsional vibration frequency of the target wind farm, thereby obtaining the control monitoring information.

[0059] After obtaining control and monitoring information, the subsynchronous torsional vibration frequency of the target thermal power unit is summed with the shaft torsional vibration frequency of the target wind farm to obtain the frequency sum; then, the pre-stored predetermined grid fundamental frequency is read and the frequency sum is compared with the predetermined grid fundamental frequency; when the frequency sum is equal to the predetermined grid fundamental frequency, or the difference between the frequency sum and the predetermined grid fundamental frequency is less than or equal to the preset frequency error threshold, it is determined that there is a risk of subsynchronous oscillation and a risk warning is issued; when the difference between the frequency sum and the predetermined grid fundamental frequency is greater than the preset frequency error threshold, no risk warning is issued and the current control state is maintained.

[0060] When retrieving the risk emergency plan based on risk warning to carry out emergency handling of the target wind farm, the risk emergency plan corresponding to the current risk status is read after the risk warning is issued, and the current inner loop gain coefficient of the rotor-side frequency converter of the target wind farm is read. The risk emergency plan refers to the adjustment of the current inner loop gain coefficient of the rotor-side frequency converter of the target wind farm based on a predetermined step size. Subsequently, the current inner loop gain coefficient is set as the first gain coefficient, and the predetermined step size is set as the gain adjustment step size. The second gain coefficient is obtained by subtracting the gain adjustment step size from the first gain coefficient, and the second gain coefficient is written into the rotor-side frequency converter as the new current inner loop gain coefficient. After one adjustment is completed, control monitoring information is acquired again, the sum of the subsynchronous torsional vibration frequency and the shaft torsional vibration frequency is recalculated, and compared with the predetermined grid fundamental frequency again. When the recalculated frequency sum is still equal to the predetermined grid fundamental frequency, or the difference between the recalculated frequency sum and the predetermined grid fundamental frequency is less than or equal to the preset frequency error threshold, the current inner loop gain coefficient is adjusted again according to the predetermined step size. When the difference between the recalculated frequency sum and the predetermined grid fundamental frequency is greater than the preset frequency error threshold, the adjustment is stopped and the currently written current inner loop gain coefficient is maintained, thereby completing the emergency handling of the target wind farm based on risk warning.

[0061] Furthermore, in the method provided in the application embodiment, after retrieving the risk emergency plan based on the risk warning to perform emergency handling on the target wind farm, it further includes: Obtain emergency monitoring information and determine whether there is a risk of subsynchronous oscillation based on the emergency monitoring information; if so, issue a power fluctuation command and perform power fluctuation on the target grid-side load based on the power fluctuation command; wherein, a subsynchronous-fluctuation lookup table is introduced, and a fluctuation strategy is determined in combination with the subsynchronous torsional vibration frequency, and the power fluctuation of the target grid-side load is performed according to the fluctuation strategy; wherein, the fluctuation strategy includes fluctuation frequency and fluctuation phase.

[0062] In this embodiment, when acquiring emergency monitoring information and determining whether there is a risk of subsynchronous oscillation based on the emergency monitoring information, after completing the emergency handling of the target wind farm based on risk warning, synchronous monitoring of the target thermal power unit and the target wind farm continues according to a preset sampling period to obtain corresponding sampling data; then, multiple consecutive sampling points are used to form an analysis time window, and the sampling data within the analysis time window are arranged in chronological order to form a time-domain data sequence, and then a discrete Fourier transform is performed on the time-domain data sequence to obtain spectrum data; then, the amplitude corresponding to each frequency point is read point by point within a preset frequency range, and the frequency point with the largest amplitude is determined as the target frequency point, wherein the target The target frequency point corresponding to the thermal power unit is determined as the subsynchronous torsional vibration frequency, and the target frequency point corresponding to the target wind farm is determined as the shaft torsional vibration frequency, thereby obtaining emergency monitoring information. After obtaining the emergency monitoring information, the subsynchronous torsional vibration frequency and the shaft torsional vibration frequency are summed to obtain the frequency sum. Then, the frequency sum is compared with the preset predetermined grid fundamental frequency. When the frequency sum is equal to the predetermined grid fundamental frequency, or the difference between the frequency sum and the predetermined grid fundamental frequency is less than or equal to the preset frequency error threshold, it is determined that there is a risk of subsynchronous oscillation. When the difference between the frequency sum and the predetermined grid fundamental frequency is greater than the preset frequency error threshold, it is determined that there is no risk of subsynchronous oscillation.

[0063] If a subsynchronous oscillation risk is identified, a power fluctuation command is issued, and power fluctuations are applied to the target grid-side load based on this command. Specifically, a subsynchronous-fluctuation lookup table is first introduced, and the subsynchronous torsional oscillation frequency is compared with each frequency interval in the table. When the subsynchronous torsional oscillation frequency falls into a certain frequency interval, the corresponding fluctuation frequency and fluctuation phase are read, and these are determined as the fluctuation strategy, which includes both fluctuation frequency and fluctuation phase. Then, the current reference power of the target grid-side load and the preset sampling period are read. Let the fluctuation frequency be f and the preset sampling period be Ts. First, the fluctuation period T is calculated by dividing the fluctuation period by 1, and then the number of sampling points N within one fluctuation period is calculated by dividing T by Ts. When N is an integer, it is directly determined as the number of sampling points within one fluctuation period; when N is not an integer, it is rounded down, and the rounded result is determined as the number of sampling points within one fluctuation period. Finally, the fluctuation phase is set as... And expressed in degrees, then according to The starting sampling point number K is calculated by dividing by 360 and then multiplying by N. When K is an integer, the sampling point corresponding to the integer is determined as the starting sampling point. When K is not an integer, K is rounded down and the sampling point corresponding to the rounded result is determined as the starting sampling point.

[0064] After determining the initial sampling point, the baseline power of the target grid-side load is used as the reference value, and the target power value corresponding to each sampling point within a fluctuation cycle is generated point by point according to the pre-set power fluctuation amplitude. Specifically, starting from the initial sampling point, N sampling points within a fluctuation cycle are numbered sequentially. When a sampling point is in the first half of the fluctuation cycle, the power fluctuation amplitude is added to the baseline power to obtain the target power value of the corresponding sampling point. When a sampling point is in the second half of the fluctuation cycle, the power fluctuation amplitude is subtracted from the baseline power to obtain the target power value of the corresponding sampling point. After completing the calculation of the target power value corresponding to each sampling point within a fluctuation cycle, the target power values ​​corresponding to each sampling point are arranged in chronological order to form a power fluctuation command. Subsequently, the target power value in the power fluctuation command is read from each sampling point, and the target power value is sent to the control terminal of the target grid-side load. When the target power value of a certain sampling point is... When the target power value is greater than the current actual power, the target grid-side load is controlled to increase its power consumption until the target power value corresponding to the sampling point is reached. When the target power value corresponding to a sampling point is less than the current actual power, the target grid-side load is controlled to decrease its power consumption until the target power value corresponding to the sampling point is reached. After completing one fluctuation cycle, emergency monitoring information is acquired again, and the risk of subsynchronous oscillation is determined in the same way. If the risk of subsynchronous oscillation is still determined, a new fluctuation strategy is determined by combining the subsynchronous-fluctuation comparison table and the power fluctuation is continued. If the risk of subsynchronous oscillation is determined, the power fluctuation command is stopped and the target grid-side load is restored to the baseline power.

[0065] In summary, the embodiments of this application have at least the following technical effects: This application acquires multi-source data, including ultra-short-term wind power forecast data of the target wind farm, real-time operation data of the target thermal power unit, and adjustable potential data of the target grid-side load. A deep-schedule cost contingency plan is introduced to analyze the cost of the real-time operation data, yielding the deep-schedule cost of the target thermal power unit. A hierarchical collaborative control architecture is constructed, comprising an upper-level optimization model and a lower-level optimization model. The upper-level optimization model, with the goal of minimizing wind curtailment, performs collaborative analysis on the ultra-short-term wind power forecast data and the adjustable potential data to obtain an upper-level scheduling decision. The lower-level optimization model, based on the upper-level scheduling decision and the deep-schedule cost, obtains the total cost and, with the goal of minimizing the total cost, obtains a lower-level scheduling decision, including the start-stop output scheme of the target thermal power unit. The target thermal power unit is controlled according to the start-stop output scheme. This invention addresses the technical problem of large wind curtailment when wind power is connected to thermal power plant substations in the prior art. By constructing a hierarchical collaborative control architecture including an upper-level optimization model and a lower-level optimization model, it coordinates the scheduling of ultra-short-term wind power forecast data, real-time operating data of thermal power units, and grid-side load adjustment potential data, thereby achieving the technical effect of reducing wind curtailment.

[0066] Example 2, based on the same inventive concept as the wind power output control method connected to a thermal power plant substation in the aforementioned examples, such as... Figure 2 As shown, this application provides a wind power output control system connected to a thermal power plant substation. The system and method embodiments in this application are based on the same inventive concept. The system includes: The data acquisition module 11 is used to acquire multi-source data, including ultra-short-term wind power forecast data of the target wind farm, real-time operation data of the target thermal power unit, and adjustable potential data of the target grid-side load. The cost analysis module 12 is used to introduce a deep adjustment cost plan to perform cost analysis on the real-time operation data to obtain the deep adjustment cost of the target thermal power unit. The architecture construction module 13 is used to construct a hierarchical collaborative control architecture, including an upper-level optimization model and a lower-level optimization model. The collaborative analysis module 14 is used for the upper-level optimization model to perform collaborative analysis on the ultra-short-term wind power forecast data and the adjustable potential data with the goal of minimizing wind curtailment, to obtain an upper-level scheduling decision. The lower-level scheduling decision determination module 15 is used for the lower-level optimization model to obtain a total cost based on the upper-level scheduling decision and the deep adjustment cost, and to obtain a lower-level scheduling decision with the goal of minimizing the total cost, including the start-stop output scheme of the target thermal power unit. The control module 16 is used to control the target thermal power unit according to the start-stop output scheme.

[0067] Furthermore, the system is also used to implement the following functions: Acquire control and monitoring information, including the subsynchronous torsional vibration frequency of the target thermal power unit and the shaft torsional vibration frequency of the target wind farm; issue a risk warning when the sum of the subsynchronous torsional vibration frequency and the shaft torsional vibration frequency reaches a predetermined grid fundamental frequency; retrieve the risk emergency plan based on the risk warning to carry out emergency handling of the target wind farm; wherein, the risk emergency plan refers to adjusting the current inner loop gain coefficient of the rotor-side frequency converter of the target wind farm based on a predetermined step size.

[0068] Furthermore, the system is also used to implement the following functions: Obtain emergency monitoring information and determine whether there is a risk of subsynchronous oscillation based on the emergency monitoring information; if so, issue a power fluctuation command and perform power fluctuation on the target grid-side load based on the power fluctuation command; wherein, a subsynchronous-fluctuation lookup table is introduced, and a fluctuation strategy is determined in combination with the subsynchronous torsional vibration frequency, and the power fluctuation of the target grid-side load is performed according to the fluctuation strategy; wherein, the fluctuation strategy includes fluctuation frequency and fluctuation phase.

[0069] Furthermore, the system is also used to implement the following functions: The real-time operating data is subjected to variation weighting calculation to obtain the real-time operating coefficient; the initial stable combustion load of the target thermal power unit is adjusted with the real-time operating coefficient as the weight to obtain the minimum stable combustion load; the target thermal power unit is divided into peak shaving categories based on the minimum stable combustion load to obtain the target category results; the target category results are analyzed for deep peak shaving cost according to the deep shaving cost plan to obtain the deep shaving cost.

[0070] Furthermore, the system is also used to implement the following functions: The target segmentation result includes a deep peak shaving stage without oil injection. The deep peak shaving cost analysis is performed on the target segmentation result according to the deep peak shaving cost plan to obtain the deep peak shaving cost, including: calculating the rotor damage cost of the target thermal power unit in the deep peak shaving stage without oil injection according to the deep peak shaving cost plan, and using the rotor damage cost as the deep peak shaving cost.

[0071] Furthermore, the system is also used to implement the following functions: The target segmentation result includes the oil injection depth peak shaving stage. According to the deep peak shaving cost plan, the target segmentation result is analyzed for deep peak shaving cost to obtain the deep peak shaving cost, including: according to the deep peak shaving cost plan, in the oil injection depth peak shaving stage, the fuel cost of the target thermal power unit is calculated; the sum of the rotor damage cost and the fuel cost is taken as the deep peak shaving cost.

[0072] Furthermore, the system is also used to implement the following functions: A tiered allocation strategy is introduced, aiming to minimize wind curtailment. This involves collaborative analysis of the ultra-short-term wind power forecast data and the adjustable potential data to determine the upper-level dispatch decision. The tiered allocation strategy includes: when a surge in wind power generation leads to insufficient peak-shaving capacity, configuring the increased power of the target grid-side load within a zero-cost adjustment range to form a first load power adjustment plan, which is then used as the upper-level dispatch decision; when the zero-cost adjustment range is exhausted, dynamically calculating the real-time adjustable capacity of the target grid-side load based on the power-temperature coupling constraint of thermal inertia to form a second load power adjustment plan, which is then used as the upper-level dispatch decision; and when both the zero-cost adjustment range and the real-time adjustable capacity are exhausted, performing deep peak shaving on the target thermal power units to form a total output plan for the thermal power units, which is then used as the upper-level dispatch decision.

[0073] Furthermore, the system is also used to implement the following functions: The maximum temperature adjustment amount is obtained based on the actual temperature of the target grid-side load and the preset safe temperature; a predetermined factor is retrieved and combined with the maximum temperature adjustment amount to determine the maximum allowable adjustment power; the second load power adjustment plan is formed with the maximum allowable adjustment power as a constraint.

[0074] Furthermore, the system is also used to implement the following functions: The total cost includes at least the coal consumption cost of the target thermal power unit, the deep adjustment cost, and the start-up and shutdown cost.

[0075] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0076] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A method for controlling wind power output connected to a thermal power plant substation, characterized in that, include: Acquire multi-source data, including ultra-short-term wind power forecast data of the target wind farm, real-time operation data of the target thermal power unit, and adjustable potential data of the target grid-side load; A deep adjustment cost contingency plan is introduced to perform cost analysis on the real-time operating data to obtain the deep adjustment cost of the target thermal power unit; A hierarchical collaborative control architecture is constructed, wherein the hierarchical collaborative control architecture includes an upper-layer optimization model and a lower-layer optimization model; The upper-level optimization model aims to minimize wind curtailment and performs collaborative analysis on the ultra-short-term wind power prediction data and the adjustable potential data to obtain upper-level scheduling decisions. The lower-level optimization model is based on the upper-level scheduling decision, obtains the total cost according to the deep adjustment cost, and obtains the lower-level scheduling decision with the goal of minimizing the total cost. The lower-level scheduling decision includes the start-up and shutdown output scheme of the target thermal power unit. The target thermal power unit is controlled according to the start-stop power output scheme.

2. The wind power output control method for connecting to a thermal power plant substation as described in claim 1, characterized in that, The target thermal power unit is controlled according to the start-stop power output scheme, and then the process further includes: Acquire control and monitoring information, wherein the control and monitoring information includes the subsynchronous torsional vibration frequency of the target thermal power unit and the shaft torsional vibration frequency of the target wind farm; When the sum of the subsynchronous torsional vibration frequency and the shaft system torsional vibration frequency reaches the predetermined power grid fundamental frequency, a risk warning is issued. Based on the aforementioned risk warning, the risk emergency plan is retrieved to carry out emergency response for the target wind farm; The aforementioned risk emergency plan refers to adjusting the inner loop gain coefficient of the rotor-side frequency converter of the target wind farm based on a predetermined step size.

3. The wind power output control method for connecting to a thermal power plant substation as described in claim 2, characterized in that, Based on the aforementioned risk warning, the risk emergency response plan is retrieved to conduct emergency handling of the target wind farm, followed by: Obtain emergency monitoring information and determine whether there is a risk of subsynchronous oscillation based on the emergency monitoring information; If present, issue a power fluctuation command and perform power fluctuation on the target grid-side load based on the power fluctuation command; The subsynchronous-fluctuation comparison table is introduced, and the fluctuation strategy is determined in combination with the subsynchronous torsional vibration frequency. The power fluctuation of the target grid-side load is then performed according to the fluctuation strategy. The fluctuation strategy includes fluctuation frequency and fluctuation phase.

4. The wind power output control method for connecting to a thermal power plant substation as described in claim 1, characterized in that, A deep adjustment cost contingency plan is introduced to perform cost analysis on the real-time operating data, resulting in the deep adjustment cost of the target thermal power unit, including: The real-time operating data is subjected to a variation-weighted calculation to obtain the real-time operating coefficients; The initial stable combustion load of the target thermal power unit is adjusted using the real-time operating coefficient as a weight to obtain the minimum stable combustion load; The target thermal power units are divided into peak-shaving categories based on the minimum stable combustion load, and the target classification results are obtained. Based on the deep adjustment cost plan, a deep peak-shaving cost analysis is performed on the target partitioning results to obtain the deep adjustment cost.

5. The wind power output control method for connecting to a thermal power plant substation as described in claim 4, characterized in that, The target segmentation result includes a deep peak shaving stage without oil injection. The deep peak shaving cost analysis is performed on the target segmentation result according to the deep peak shaving cost plan to obtain the deep peak shaving cost, including: calculating the rotor damage cost of the target thermal power unit in the deep peak shaving stage without oil injection according to the deep peak shaving cost plan, and using the rotor damage cost as the deep peak shaving cost.

6. The wind power output control method for connecting to a thermal power plant substation as described in claim 5, characterized in that, The target classification result includes the oil injection depth peak shaving stage. According to the deep peak shaving cost plan, the target classification result is analyzed for deep peak shaving cost to obtain the deep peak shaving cost, including: according to the deep peak shaving cost plan, in the oil injection depth peak shaving stage, the fuel cost of the target thermal power unit is calculated. The sum of the rotor wear cost and the fuel cost is taken as the deep adjustment cost.

7. The wind power output control method for connecting to a thermal power plant substation as described in claim 1, characterized in that, The upper-level optimization model aims to minimize wind curtailment. It performs collaborative analysis on the ultra-short-term wind power forecast data and the adjustable potential data to obtain upper-level scheduling decisions, including: A tiered allocation strategy is introduced, aiming to minimize wind curtailment. This involves collaborative analysis of the ultra-short-term wind power forecast data and the adjustable potential data to determine the upper-level scheduling decision. The tiered allocation strategy includes: When wind power generation leads to insufficient peak-shaving capacity, the upward adjustment power of the target grid-side load is configured in the zero-cost adjustment range to form a first load power adjustment plan, and the first load power adjustment plan is used as the upper-level dispatch decision. Once the zero-cost adjustment range is exhausted, the real-time adjustable capacity of the target grid-side load is dynamically calculated based on the power-temperature coupling constraint of thermal inertia, forming a second load power adjustment plan, and the second load power adjustment plan is used as the upper-level scheduling decision. Once the zero-cost adjustment range and real-time adjustable capacity are exhausted, deep peak shaving is performed on the target thermal power unit to form a total output plan for the thermal power unit, and the total output plan for the thermal power unit is used as the upper-level scheduling decision.

8. The wind power output control method for connecting to a thermal power plant substation as described in claim 7, characterized in that, Once the zero-cost adjustment range is exhausted, the real-time adjustable capacity of the target grid-side load is dynamically calculated based on the power-temperature coupling constraint of thermal inertia, forming a second load power adjustment plan, including: The maximum temperature adjustment amount is obtained based on the actual temperature of the target grid-side load and the preset safe temperature. The predetermined factor is retrieved and combined with the maximum temperature adjustment amount to determine the maximum allowable adjustment power; The second load power regulation plan is formed with the maximum allowable adjustment power as a constraint.

9. The wind power output control method for connecting to a thermal power plant substation as described in claim 1, characterized in that, The total cost includes at least the coal consumption cost of the target thermal power unit, the deep adjustment cost, and the start-up and shutdown cost.

10. A wind power output control system connected to a thermal power plant substation, characterized in that, The system is used to execute the wind power output control method for connecting to a thermal power plant substation as described in any one of claims 1-9, and the system includes: The data acquisition module is used to acquire multi-source data, including ultra-short-term wind power forecast data of the target wind farm, real-time operation data of the target thermal power unit, and adjustable potential data of the target grid-side load. The cost analysis module is used to introduce a deep adjustment cost plan to perform cost analysis on the real-time operating data and obtain the deep adjustment cost of the target thermal power unit. An architecture building module is used to build a hierarchical collaborative control architecture, wherein the hierarchical collaborative control architecture includes an upper-layer optimization model and a lower-layer optimization model; The collaborative analysis module is used by the upper-level optimization model to perform collaborative analysis on the ultra-short-term wind power prediction data and the adjustable potential data with the goal of minimizing wind curtailment, so as to obtain upper-level scheduling decisions. The lower-level scheduling decision determination module is used by the lower-level optimization model to obtain the total cost based on the upper-level scheduling decision and the deep adjustment cost, and to obtain the lower-level scheduling decision with the goal of minimizing the total cost. The lower-level scheduling decision includes the start-up and shutdown output scheme of the target thermal power unit. The control module is used to control the target thermal power unit according to the start-stop power output scheme.

Images