Coal-fired power generation unit operation optimization method and system based on industrial big data

Abstract

The invention discloses a coal-fired power generation unit operation optimization method and system based on industrial big data, and relates to the technical field of industries. A starting priority analysis module dynamically adjusts a unit starting sequence and obtains unit starting priorities by analyzing real-time data sets and power grid load requirements; the system can quickly judge the starting sequence of the units according to the numerical values and effectively recognize the units with possible fault risks, and the fault recognition module performs fault assessment on the unit operation state and equipment state data in combination with the real-time data set and the historical data set and calculates the current fault assessment index. Through comprehensive analysis of a historical trend and a real-time state, the system can accurately judge whether maintenance operation needs to be carried out on a unit or not, so that early prediction and timely intervention of potential faults are realized, an effect feedback and optimization module evaluates feedback data of an actual operation effect, constructs a performance feedback index, and improves the maintenance efficiency of the unit. And dynamic adjustment and optimization of unit load distribution are realized.

Description

Technical Field

[0001] The present invention relates to the field of industrial technology, and in particular to a method and system for optimizing the operation of a coal-fired power generation unit based on industrial big data. Background Art

[0002] As an important part of global power production, coal-fired power generation units are widely used in the field of traditional energy and are the core of stable operation of power grids. With the continuous advancement of industrial technology, industrial big data has become one of the key technologies to promote energy management and optimization, injecting new vitality into the traditional energy industry through real-time monitoring, data analysis and intelligent optimization. In the field of combining industrial big data with coal-fired power generation units, intelligent optimization of the operation of power generation units based on a large amount of real-time and historical data has become one of the hot spots of academic research and industrial application. Specifically, according to the real-time load data of the power grid and the operating status of the coal-fired power generation units, optimizing the start-stop sequence and load distribution of the units can significantly improve the load response capability of the power grid, reduce energy consumption, and improve the economic efficiency of the unit operation. This big data-based optimization technology effectively integrates data analysis and system scheduling, providing a solution for the intelligent, green and efficient coal-fired power generation.

[0003] In the current operation and management of coal-fired power generation units, traditional start-stop sequences and load distribution strategies often rely on manual experience or simple rules, and lack the ability to dynamically adjust based on real-time data. This management model may lead to response delays when load demand fluctuates greatly, and it is difficult to quickly allocate resources to meet grid demand. In addition, due to the lack of in-depth analysis of the historical operation data of the unit, the economy and availability of the unit are difficult to be optimally guaranteed under complex operating conditions, resulting in excessive fuel consumption and reduced power generation efficiency. Especially for potential unit failures, traditional methods rely on regular inspections or equipment alarms, which makes it difficult to predict trends and intervene early, often leading to delayed or over-maintenance of equipment maintenance. Therefore, the current system has significant deficiencies in optimization efficiency, grid load response capability, and fault prediction capability, and an intelligent optimization system based on industrial big data is urgently needed to improve it. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention provides a method and system for optimizing the operation of a coal-fired power generation unit based on industrial big data, which solves the problems in the above-mentioned background technology.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: an operation optimization system for coal-fired power generation units based on industrial big data, including a data acquisition module, a data processing module, a startup priority analysis module, a fault identification module, and an effect feedback and optimization module;

[0006] The data acquisition module is used to use multiple groups of monitoring instruments to monitor the relevant unit operation data and relevant equipment status data of each coal-fired power generation unit in real time, generate a real-time data group, and obtain the historical data group of each coal-fired power generation unit in a historical period based on big data technology;

[0007] The data processing module is used to perform data preprocessing on the relevant data in the real-time data group, remove noise, eliminate outliers, and fill in missing values, and after dimensionless processing, construct a processed real-time data group;

[0008] The startup priority analysis module is used to analyze and determine the startup sequence of each unit based on the processed real-time data group's related unit operation data and combined with the power grid load demand Dfyq to obtain the unit startup priority Qyxj. According to the numerical value of the unit startup priority Qyxj, the startup sequence of each unit and whether there is a failure risk are determined. If there is a failure risk, the trend analysis instruction is triggered;

[0009] The fault identification module is used to receive trend analysis instructions and analyze the current fault assessment index Ggzs of each unit during operation according to the relevant equipment status data in the processed real-time data group. c , and combined with the historical data group, determine whether the corresponding unit needs maintenance work;

[0010] The effect feedback and optimization module is used to provide feedback based on the actual effect formed by the startup sequence of each unit, obtain relevant feedback data, and evaluate the balance effect between the grid load response capability and the unit operating efficiency based on the relevant feedback data to construct a performance feedback index Xfzs. Based on the performance feedback index Xfzs, the load of the corresponding unit is adjusted and optimized.

[0011] Preferably, the data acquisition module includes a real-time monitoring unit and a history extraction unit;

[0012] The real-time monitoring unit is used to monitor the relevant unit operation data and relevant equipment status data of each coal-fired power generation unit in real time using multiple sets of monitoring instruments, wherein the relevant unit operation data includes the unit time fuel consumption Rxz, load output value Fsz, availability coefficient Kyz of the corresponding unit at each time, unit load fluctuation factor Fbyz and maximum power generation load output value Fbz of the corresponding unit max ; The relevant equipment status data includes boiler pressure fluctuation factor Ybyz1 and steam pipeline pressure fluctuation factor Ybyz1; multiple groups of monitoring instruments include laser coal flow meter, power monitoring instrument and unit performance test instrument; the real-time data group includes relevant unit operation data and relevant equipment status data;

[0013] The history extraction unit is used to obtain relevant unit operation data and relevant equipment status data of each coal-fired power generation unit in the industrial park within a historical period using big data technology. The historical data group includes relevant unit operation data and relevant equipment status data within the historical period.

[0014] Preferably, the data processing module includes a preprocessing unit and a normalization unit;

[0015] The preprocessing unit is used to preprocess the relevant data in the real-time data group, and the preprocessing includes removing noise, filling missing values, removing outliers and data smoothing operations, wherein the missing value filling methods include mean filling, median filling, interpolation filling and regression filling;

[0016] The normalization unit is used to eliminate unit differences of relevant data in the preprocessed real-time data group by using dimensionless processing technology, so that the range of relevant data in the preprocessed real-time data group falls within [0, 1].

[0017] Preferably, the startup priority analysis module includes an operation status analysis unit, a priority unit and a preliminary judgment unit;

[0018] The operation status analysis unit is used to extract the unit time fuel consumption Rxz and the load output value Fsz of each unit from the processed real-time data group of the relevant unit operation data, and calculate the economic index Jzb of each unit through the unit time fuel consumption Rxz and the load output value Fsz of each unit. The economic index Jzb of each unit is obtained specifically according to the following formula:

[0019]

[0020] According to the economic index Jzb of each unit and the load demand of the power grid Dfyq, the target value of the power generation load Fbz of each unit in different monitoring periods is calculated. Specifically, the target value of the power generation load Fbz of each unit in different monitoring periods is obtained according to the following formula:

[0021]

[0022] In the formula, Fbz(t) represents the target power generation load value of the corresponding unit at time t, Dfyq(t) represents the grid load demand at time t, and Jzb -1 represents the economic weight of the corresponding unit; n represents the number of units, i = 1, 2, 3, ..., n.

[0023] Preferably, the priority unit is used to analyze and determine the startup sequence of each unit based on the processed relevant unit operation data in the real-time data group and in combination with the current grid load demand to obtain the unit startup priority Qyxj, and the unit startup priority Qyxj is specifically obtained by the following formula:

[0024]

[0025] Where Qyxj(t) represents the unit startup priority of the corresponding unit at time t, Kyz(t) represents the availability coefficient of the corresponding unit at time t, and Fbz max It is represented by the maximum power generation load output value of the corresponding unit, and ∈ is represented by the correction constant.

[0026] Preferably, the preliminary judgment unit is used to obtain the unit startup priority Qyxj of each unit according to the method of obtaining the unit startup priority Qyxj in the priority unit, and sort them to generate a sequence group; determine the unit startup priority Qyxj of the corresponding unit in the historical period according to the historical data group, and obtain the average unit startup priority Qyxj of the corresponding unit in combination with the statistical averaging algorithm. avg , by comparing the unit startup priority Qyxj of the current unit with the average unit startup priority Qyxj of the corresponding unit avg Compare to determine the startup sequence of the corresponding units and whether there is a risk of failure. The specific preliminary judgment is as follows:

[0027] If the unit startup priority Qyxj of the current unit ≥ the average unit startup priority Qyxj of the corresponding unit avg When the corresponding unit is initially determined to have no failure risk, the trend analysis command is not triggered externally, the corresponding unit is retained in the sequence group, and the corresponding unit is marked as a normal unit;

[0028] If the unit startup priority Qyxj of the current unit is less than the average unit startup priority Qyxj of the corresponding unit avg When the corresponding unit is initially determined to have a failure risk, a trend analysis instruction is triggered externally to remove the corresponding unit from the sequence group and mark the corresponding unit as an abnormal unit;

[0029] According to the preliminary judgment, the normal units are counted and re-sorted to generate an optimized sequence group. Based on the optimized sequence group, the startup order of each normal unit is determined.

[0030] Preferably, the fault identification module includes a fault analysis unit and an identification unit;

[0031] The fault analysis unit is used to analyze the current fault assessment index Ggzs of each abnormal unit during operation according to the processed relevant equipment status data in the real-time data group after receiving the trend analysis instruction and after dimensionless processing. c , the current fault assessment index Ggzs c Obtained by the following formula:

[0032]

[0033] Where ΔQyxj represents the unit startup priority difference, Ybyz1 represents the boiler pressure fluctuation factor, Ybyz2 represents the steam pipeline pressure fluctuation factor, Fbyz represents the unit load fluctuation factor, α, β, Y and All represent weight values, where 0<α<1, 0<β<1, 0<Y<1, α, β, γ and The specific value is set by the user according to the situation;

[0034] The identification unit is used to determine the time point when the current corresponding abnormal unit was marked as an abnormal unit last time, marking it as the starting timestamp, and marking the current time as the ending timestamp, based on the starting timestamp and the ending timestamp, obtain the time interval, use the time interval as the comparison period, and extract the fault assessment index Ggzs at the starting timestamp from the historical data group according to the starting timestamp h By comparing the current fault assessment index Ggzsc of each unit during operation with the fault assessment index Ggzs at the start timestamp h Compare to determine whether the corresponding abnormal unit needs maintenance. The specific contents are as follows:

[0035] If the current fault assessment index Ggzs of each unit during operation c The fault assessment index Ggzs exceeding the start timestamp h When the maintenance order is issued, maintenance personnel will be arranged to carry out on-site maintenance work;

[0036] If the current fault assessment index Ggzs of each unit during operation c The fault assessment index Ggzs at the start timestamp is not exceeded h At this time, no maintenance instruction will be issued.

[0037] Preferably, the effect feedback and optimization module includes an effect feedback unit and an optimization unit;

[0038] The effect feedback unit is used to provide feedback according to the actual effect formed by the startup sequence of each unit, and obtain relevant feedback data, wherein the relevant feedback data includes the number m of normal units, the load output value of the corresponding normal units at each moment, the power generation load target value of the corresponding normal units at each moment, and the fuel consumption of the corresponding normal units at each moment; through the relevant feedback data, the balance effect between the load response capability of the power grid and the operating efficiency of the units is evaluated, and after dimensionless processing, a performance feedback index Xfzs is constructed, and the performance feedback index Xfzs is obtained by the following formula:

[0039]

[0040] Where k = 1, 2, ..., m, m represents the number of normal units, Xfzs(t) represents the performance feedback index at time t, and Fsz k (t) represents the load output value of the kth normal unit at time t, Fbz k (t) represents the target power generation load value of the kth normal unit at time t, Rxz k (t) represents the fuel consumption of the kth normal unit at time t, and ω represents the weight coefficient.

[0041] Preferably, the optimization unit is used to use a gradient descent method and, in combination with a performance feedback index Xfzs, to adjust and optimize the load of the corresponding unit to obtain a load adjustment amount Ft of the corresponding unit at each subsequent moment:

[0042]

[0043] Where Ft(t) represents the performance feedback index at time t, and η represents the learning rate; Represents the partial derivative of the performance feedback index with respect to the load output value of the unit at time t.

[0044] A method for optimizing the operation of a coal-fired power generation unit based on industrial big data comprises the following steps:

[0045] S 1. Use multiple sets of monitoring instruments to monitor the relevant unit operation data and relevant equipment status data of each coal-fired power generation unit in real time, generate a real-time data set, and obtain the historical data set of each coal-fired power generation unit in the historical period based on big data technology;

[0046] S2, preprocessing the relevant data in the real-time data group, removing noise, eliminating outliers, and filling in missing values, and constructing a processed real-time data group after dimensionless processing;

[0047] S3. Analyze and judge the current startup sequence of each unit based on the processed real-time data group's relevant unit operation data and combined with the grid load demand Dfyq to obtain the unit startup priority Qyxj. According to the numerical value of the unit startup priority Qyxj, determine the startup sequence of each unit and whether there is a failure risk. If there is a failure risk, trigger the trend analysis instruction.

[0048] S4, receiving trend analysis instructions, analyzing the current fault assessment index Ggzs of each unit during operation according to the relevant equipment status data in the processed real-time data group c , and combined with the historical data group, determine whether the corresponding unit needs maintenance work;

[0049] S5. Provide feedback based on the actual effect of the startup sequence of each unit, obtain relevant feedback data, and evaluate the balance between the grid load response capability and the unit operating efficiency based on the relevant feedback data to construct a performance feedback index Xfzs. Based on the performance feedback index Xfzs, adjust and optimize the load of the corresponding unit.

[0050] The present invention provides a method and system for optimizing the operation of a coal-fired power generation unit based on industrial big data, which has the following beneficial effects:

[0051] (1) Through the multiple monitoring instruments of the data acquisition module, the system can collect the operating status and equipment data of the unit in real time to form a real-time data group; at the same time, combined with the historical data group, the long-term operating status of the unit is analyzed based on big data technology, and the operating rules and load characteristics of the unit are fully understood. This combination of real-time and historical data not only improves the integrity of data analysis, but also provides reliable data support for the accuracy of subsequent optimization strategies. The startup priority analysis module dynamically adjusts the unit startup sequence and obtains the unit startup priority by analyzing the real-time data group and the grid load demand. The system can quickly determine the unit startup sequence based on the numerical value and effectively identify the unit that may have a risk of failure, and further trigger the trend analysis instruction. This module reduces the blindness of the start-up and shutdown decision-making while ensuring the grid load response capability, and improves the economic efficiency of the unit operation. The fault identification module combines the real-time data group and the historical data group to perform fault assessment on the unit operation status and equipment status data, and calculates the current fault assessment index. By comprehensively analyzing historical trends and real-time status, the system can accurately determine whether maintenance work is needed on the unit, thereby achieving early prediction and timely intervention of potential faults. This function effectively reduces the risk of unplanned downtime of the unit and ensures the stability of system operation. The effect feedback and optimization module evaluates the feedback data of the actual operating effect, constructs a performance feedback index, and realizes dynamic adjustment and optimization of the unit load distribution. On this basis, the system can continuously optimize the balance between the grid load response capability and the unit operating efficiency, ensuring that the unit operates with the lowest fuel consumption while meeting the grid demand. This closed-loop optimization mechanism enables the system to continuously adjust the optimization strategy according to the actual effect, thereby achieving continuous improvement in operating efficiency.

[0052] (2) The priority unit dynamically calculates the startup priority Qyxj of each unit by combining the unit operation status and grid load demand in the real-time data group, comprehensively considering the unit's availability coefficient, maximum power generation load output value and correction constant to ensure the accuracy and robustness of the calculation. This module can quickly adjust the start-stop sequence of the unit according to the real-time grid load demand, give priority to starting units with high availability and strong power generation capacity, and effectively improve the load response capability of the grid. At the same time, it further avoids the waste of resources and economic decline caused by the participation of low-priority units in operation. The preliminary judgment unit uses the statistical mean algorithm to calculate the average startup priority of each unit in combination with the historical data group, and further compares and verifies the currently calculated unit startup priority Qyxj. If the current priority Qyxj is lower than the historical mean, the system can preliminarily judge that the unit has a fault risk, and promptly trigger the trend analysis command to remove the unit with possible faults from the sequence group. This mechanism realizes the accurate identification of potential faults by dynamically comparing historical and current data, improves the efficiency of fault detection, avoids the participation of faulty units in grid operation, and ensures system stability. After sorting the unit priorities Qyxj, the preliminary judgment unit generates a sequence group, eliminates potential faulty units, optimizes the unit sequence, and redetermines the start-up order of normal units based on the optimized sequence group to ensure that the system gives priority to units with good operating status and strong power generation capacity.

[0053] (3) After receiving the trend analysis instruction, the fault analysis unit combines the relevant equipment status data in the processed real-time data group and calculates the fault assessment index of each abnormal unit after dimensionless processing. Through the accurate calculation of the fault assessment index, the system can quantify the abnormal operation degree of the unit in multiple dimensions, significantly improve the accuracy of fault assessment, and provide a scientific basis for subsequent fault identification and decision-making. The identification unit accurately records the duration of the abnormal state by marking the start and end timestamps of the abnormal unit, and dynamically analyzes the development trend of the fault with the time interval as the comparison period. By extracting the fault assessment index at the start timestamp from the historical data group and comparing it with the current fault assessment index, the system can dynamically judge the severity and development trend of the fault: if the current fault assessment index exceeds the assessment value at the start timestamp, it indicates that the fault is in an aggravating trend. The system immediately triggers the maintenance instruction and arranges on-site maintenance operations to ensure that the problem is handled in a timely manner. If the current fault assessment index does not exceed the historical value, it indicates that the fault is in a controllable state. The system will not issue a maintenance instruction for the time being, avoiding unnecessary maintenance operations and reducing maintenance costs. The fault identification module effectively avoids resource waste and fault escalation caused by excessive maintenance or delayed maintenance through linkage analysis of real-time and historical data.

[0054] (4) The construction of the performance feedback index Xfzs achieves a quantitative evaluation of the grid load response capability and the unit operating efficiency through a comprehensive analysis of load error and fuel economy, greatly improving the operability and evaluation accuracy of the system. Combined with the dynamic optimization unit of the gradient descent method, the system can reduce the impact of grid load fluctuations on unit operation through a scientific load distribution strategy, reduce fuel waste, and improve power generation efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 This is a block diagram of a coal-fired power generation unit operation optimization system based on industrial big data of the present invention;

[0056] Figure 2 The present invention is a schematic flow chart of a method for optimizing the operation of a coal-fired power generation unit based on industrial big data. DETAILED DESCRIPTION

[0057] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0058] Example 1

[0059] See also Figure 1 , the present invention provides a coal-fired power generation unit operation optimization system based on industrial big data, including a data acquisition module, a data processing module, a startup priority analysis module, a fault identification module and an effect feedback and optimization module;

[0060] The data acquisition module is used to use multiple groups of monitoring instruments to monitor the relevant unit operation data and relevant equipment status data of each coal-fired power generation unit in real time, generate a real-time data group, and obtain the historical data group of each coal-fired power generation unit in a historical period based on big data technology;

[0061] The data processing module is used to perform data preprocessing on the relevant data in the real-time data group, remove noise, eliminate outliers, and fill in missing values, and after dimensionless processing, construct a processed real-time data group;

[0062] The startup priority analysis module is used to analyze and determine the startup sequence of each unit based on the processed real-time data group's related unit operation data and combined with the power grid load demand Dfyq to obtain the unit startup priority Qyxj. According to the numerical value of the unit startup priority Qyxj, the startup sequence of each unit and whether there is a fault risk are determined. If there is a fault risk, the trend analysis instruction is triggered;

[0063] The fault identification module is used to receive trend analysis instructions and analyze the current fault assessment index Ggzs of each unit during operation according to the processed relevant equipment status data in the real-time data group. c , and combined with the historical data group, determine whether the corresponding unit needs maintenance work;

[0064] The effect feedback and optimization module is used to provide feedback based on the actual effect formed by the startup sequence of each unit, obtain relevant feedback data, and evaluate the balance effect between the grid load response capability and the unit operating efficiency based on the relevant feedback data to construct a performance feedback index Xfzs. Based on the performance feedback index Xfzs, the load of the corresponding unit is adjusted and optimized.

[0065] During the operation of this system, the data acquisition module monitors the unit operation data and equipment status data in real time, and combined with the mining and analysis of historical data groups, it can grasp the dynamic changes of the power grid load demand Dfyq and the unit status. Based on the startup priority analysis module, the system can dynamically calculate the startup priority Qyxj of the computer group, and perform relatively optimal start-stop scheduling of the unit based on its numerical value, thereby improving the unit's responsiveness to fluctuations in power grid load demand and avoiding the delay problem in traditional start-stop scheduling, thereby effectively ensuring the stability of power grid operation. The real-time data group is preprocessed by the data processing module (including denoising, removing outliers, filling missing values ​​and dimensionless operations) to ensure data quality and provide accurate input for subsequent optimization analysis. Combined with the processed real-time data group and the power grid load demand, the startup priority analysis module can prioritize the scheduling of units with high operating efficiency and good economy. By accurately calculating the load distribution of the unit, energy waste and unnecessary fuel consumption are avoided, thereby reducing the overall operating cost of the unit and improving the power generation efficiency. The fault identification module uses the trend analysis instructions and the equipment status data in the real-time data group and the historical data group to calculate the fault evaluation index during the operation of the unit, and predict potential risks in combination with the historical operation trend. This module can not only accurately determine whether the unit needs to be repaired, but also help the operation and maintenance personnel to formulate maintenance plans in advance to avoid unplanned shutdowns of the unit due to sudden failures. Through intelligent fault identification and prediction, the availability of the unit is greatly improved, and downtime and maintenance costs are reduced. The effect feedback and optimization module dynamically evaluates the balance between the power grid load response capability and the unit operation efficiency by constructing the performance feedback index Xfzs based on the actual operation effect of the start-up sequence of each unit. The system can adjust and optimize the load distribution of the corresponding unit according to the feedback data, form a closed-loop control process, and realize the continuous improvement of start-stop scheduling and load distribution. In short, this method significantly improves the operating efficiency and economy of coal-fired power generation units through the in-depth application of industrial big data technology, combined with intelligent start-stop priority analysis and fault prediction mechanism, and enhances the load response capability of the power grid, while reducing the economic losses caused by fault shutdown. This optimization system has injected intelligent and green concepts into the traditional coal-fired power generation industry, and provided strong technical support for energy management and sustainable development.

[0066] Example 2

[0067] Please refer to Figure 1 ,Specifically: the data acquisition module includes a real-time monitoring unit and a history extraction unit;

[0068] The real-time monitoring unit is used to monitor the relevant unit operation data and relevant equipment status data of each coal-fired power generation unit in real time using multiple sets of monitoring instruments, wherein the relevant unit operation data includes the unit time fuel consumption Rxz, load output value Fsz, availability coefficient Kyz of the corresponding unit at each time, unit load fluctuation factor Fbyz and maximum power generation load output value Fbz of the corresponding unit max ; The relevant equipment status data includes boiler pressure fluctuation factor Ybyz1 and steam pipeline pressure fluctuation factor Ybyz1; multiple groups of monitoring instruments include laser coal flow meter, power monitoring instrument and unit performance test instrument; the real-time data group includes relevant unit operation data and relevant equipment status data;

[0069] The history extraction unit is used to obtain relevant unit operation data and relevant equipment status data of each coal-fired power generation unit in the industrial park within a historical period using big data technology. The historical data group includes relevant unit operation data and relevant equipment status data within the historical period.

[0070] The data processing module includes a preprocessing unit and a normalization unit;

[0071] The preprocessing unit is used to preprocess the relevant data in the real-time data group, and the preprocessing includes removing noise, filling missing values, removing outliers and data smoothing operations, wherein the missing value filling methods include mean filling, median filling, interpolation filling and regression filling;

[0072] The normalization unit is used to eliminate unit differences of relevant data in the preprocessed real-time data group by using dimensionless processing technology, so that the range of relevant data in the preprocessed real-time data group falls within [0, 1].

[0073] In this embodiment, the real-time monitoring unit in the data acquisition module monitors the unit operation data and equipment status data in real time through multiple groups of monitoring instruments (including boilers, steam pipes, etc.) to generate real-time data groups. These real-time data can fully reflect the immediate operation status and equipment health status of the unit. At the same time, the historical extraction unit extracts the historical operation data and equipment status data of each unit in the industrial park through big data technology to form a historical data group. This combination of real-time and historical data provides a multi-dimensional reference basis for analyzing the operation rules of the unit, predicting trends, and optimizing operation strategies, and improves the depth and comprehensiveness of data analysis. The preprocessing unit in the data processing module significantly improves the quality and integrity of real-time data through denoising, removing outliers, data smoothing, and missing value filling operations. Specifically: denoising effectively eliminates random noise interference in the monitoring process and ensures the accuracy of the data. Outlier removal avoids the influence of extreme values ​​on data analysis and optimization results. Missing value filling provides a variety of filling methods (mean filling, median filling, interpolation filling, and regression filling), which flexibly adapts to the processing requirements of different data types and further ensures the integrity of the data. Data smoothing eliminates the sharp fluctuations in data, improves the continuity and trend of data, and provides stable basic data for subsequent optimization analysis. The normalization unit in the data processing module eliminates the unit differences of the pre-processed real-time data group through dimensionless processing technology, so that data of different dimensions can be unified into the same range. For example: after the unit time fuel consumption and load output value are dimensionlessly processed, they can be compared and optimized under the same standard. This processing process eliminates the limitations of physical units on data analysis and significantly improves the comparability of data and the accuracy of model analysis. Through the combination of real-time monitoring units and historical extraction units, the system can conduct comprehensive real-time monitoring of the operating status and equipment status of coal-fired power generation units, and conduct in-depth analysis combined with historical data, thereby providing data support for optimizing unit operation.

[0074] Example 3

[0075] Please refer to Figure 1 ,Specifically: the startup priority analysis module includes an operation status analysis unit, a priority unit and a preliminary judgment unit;

[0076] The operation status analysis unit is used to extract the unit time fuel consumption Rxz and the load output value Fsz of each unit from the processed real-time data group of the relevant unit operation data, and calculate the economic index Jzb of each unit through the unit time fuel consumption Rxz and the load output value Fsz of each unit. The economic index Jzb of each unit is obtained specifically according to the following formula:

[0077]

[0078] The economic index Jzb of each unit refers to the unit's power generation cost per unit time (usually in yuan / kWh), which reflects the economic efficiency of the unit's operation, specifically the fuel and operating cost required for the unit to produce 1 kWh of electricity. A lower economic index Jzb means that the unit has a lower operating cost and better economy, while a lower economic index Jzb means that the unit has a higher operating cost and poorer economy.

[0079] The unit time fuel consumption Rxz of each unit can be monitored and obtained by a laser coal flow meter;

[0080] The load output value Fsz refers to the generating power of the unit, which can be monitored and obtained through the power monitoring instrument;

[0081] According to the economic index Jzb of each unit and the load demand of the power grid Dfyq, the target value of the power generation load Fbz of each unit in different monitoring periods is calculated. Specifically, the target value of the power generation load Fb of each unit in different monitoring periods is obtained according to the following formula: z :

[0082]

[0083] In the formula, Fbz(t) represents the target power generation load value of the corresponding unit at time t, Dfyq(t) represents the grid load demand at time t, and Jzb -1 It represents the economic weight of the corresponding unit, that is, the reciprocal of the economic index Jzb of the corresponding unit; n represents the number of units, i=1, 2, 3, ..., n.

[0084] The grid load demand Dfyq at each moment monitors the real-time load demand of the entire grid through the grid monitoring system, including the changing trend of load demand, peak and valley loads, etc.

[0085] In this embodiment, the operation status analysis unit in the startup priority analysis module can extract the unit time fuel consumption and load output value of each unit from the real-time data group, and calculate the economic index, that is, the unit time power generation cost, through the two. The economic index accurately quantifies the operating cost of each unit and reflects the economic differences between different units. The system achieves the scientific sorting of the start-stop priority of the unit by giving priority to units with lower operating costs and better economy, and reduces the adverse effects of blind start-stop on fuel consumption and operating efficiency. The system calculates the power generation load target value of each unit according to the formula by analyzing the economic index and the real-time load demand of the power grid to ensure that the power grid load demand is accurately allocated. Specifically, the system allocates load according to the economic weight of each unit. Units with lower economic indicators (lower operating costs) will be allocated higher power generation load target values ​​first. This load allocation mechanism significantly improves the overall economy of the power generation system, reduces fuel consumption and power generation costs, and optimizes the dynamic response capability of the power grid load. The allocation optimization of multi-unit load targets not only meets the power grid demand, but also avoids the problem of excessive load or idle operation of individual units, and improves the utilization efficiency of unit resources.

[0086] Example 4

[0087] Please refer to Figure 1 Specifically: the priority unit is used to analyze and determine the startup sequence of each unit according to the processed real-time data group related unit operation data and the current grid load demand, so as to obtain the unit startup priority Qyxj. The unit startup priority Qyxj is specifically obtained by the following formula:

[0088]

[0089] Where Qyxj(t) represents the unit startup priority of the corresponding unit at time t, Kyz(t) represents the availability coefficient of the corresponding unit at time t, and Fbz max It is represented by the maximum power generation load output value of the corresponding unit, ∈ is represented by the correction constant, which is a constant to avoid division by zero error;

[0090] The maximum power generation load output value Fbz of the corresponding unit max The unit performance test instrument can be used to test the unit's output under rated conditions to obtain the unit's maximum load output value.

[0091] Among them, the availability coefficient Kyz(t) of the corresponding unit at time t is determined according to the operating status of the unit, and the operating status of the unit includes whether it is under maintenance, whether a fault occurs, and whether it is in a standby state; when the unit is in a standby state, the corresponding availability coefficient = 1; if the unit is not in a standby state, the corresponding availability coefficient = 0.

[0092] The preliminary judgment unit is used to obtain the unit startup priority Qyxj of each unit according to the method of obtaining the unit startup priority Qyxj in the priority unit, and sort them to generate a sequence group; determine the unit startup priority Qyxj of the corresponding unit in the historical period according to the historical data group, and obtain the average unit startup priority Qyxj of the corresponding unit in combination with the statistical averaging algorithm avg , by comparing the unit startup priority Qyxj of the current unit with the average unit startup priority Qyxj of the corresponding unit avg Compare to determine the startup sequence of the corresponding units and whether there is a risk of failure. The specific preliminary judgment is as follows:

[0093] If the unit startup priority Qyxj of the current unit ≥ the average unit startup priority Qyxj of the corresponding unit avg When the corresponding unit is initially determined to have no failure risk, the trend analysis command is not triggered externally, the corresponding unit is retained in the sequence group, and the corresponding unit is marked as a normal unit;

[0094] If the unit startup priority Qyxj of the current unit is less than the average unit startup priority Qyxj of the corresponding unit avg When the corresponding unit is initially determined to have a failure risk, a trend analysis instruction is triggered externally to remove the corresponding unit from the sequence group and mark the corresponding unit as an abnormal unit;

[0095] According to the preliminary judgment, the normal units are counted and re-sorted to generate an optimized sequence group. Based on the optimized sequence group, the startup order of each normal unit is determined.

[0096] In this embodiment, the priority unit dynamically calculates the unit startup priority Qyxj of each unit by combining the key parameters such as the availability coefficient and the maximum load output value of the unit in the real-time data group with the load demand of the power grid. This calculation method ensures that the unit will give priority to the units with good startup status and high operating economy while meeting the load demand of the power grid. Through the statistical analysis of the historical data group, the system obtains the average startup priority of each unit, and uses the comparison of the current priority with the historical average priority to accurately identify whether the current startup status of the unit meets the expectations, thereby further optimizing the startup sequence of the unit. The preliminary judgment unit calculates the startup priority Qyxj of the computer group in real time and compares it with the historical average priority to quickly identify the abnormal state of the unit: when the startup priority Qyxj is lower than the average startup priority, it is preliminarily determined that the unit has a risk of failure, and the trend analysis instruction is triggered in time to further confirm the source of the problem; when the startup priority Qyxj is higher than or equal to the average startup priority, the unit is judged to be in normal state, avoiding unnecessary maintenance or interruption. This mechanism realizes early warning of potential faults, effectively reduces the risk of unplanned equipment downtime, and extends the operating life of the equipment. Based on the preliminary judgment results, the system automatically removes abnormal units and reorders normal units to generate an optimized unit startup sequence. The optimized sequence group further improves the load response capability of the power grid, ensures that units with good economic performance and stable status are started first, avoids the operation of inefficient units or units with risks, and reduces fuel consumption and operating costs.

[0097] Example 5

[0098] Please refer to Figure 1 ,Specifically: the fault identification module includes a fault analysis unit and an identification unit;

[0099] The fault analysis unit is used to analyze the current fault assessment index Ggzs of each abnormal unit during operation according to the processed relevant equipment status data in the real-time data group after receiving the trend analysis instruction and after dimensionless processing. c , the current fault assessment index Ggzs c Obtained by the following formula:

[0100]

[0101] Where ΔQyxj represents the unit startup priority difference, Ybyz1 represents the boiler pressure fluctuation factor, Ybyz2 represents the steam pipeline pressure fluctuation factor, Fbyz represents the unit load fluctuation factor, α, β, Y and All represent weight values, where 0<α<1, 0<β<1, 0<Y<1, α, β, γ and The specific value is set by the user according to the situation;

[0102] The above-mentioned unit startup priority difference ΔQyxj refers to the difference between the startup priority of the current corresponding unit and the startup priority obtained by the previous monitoring;

[0103] The boiler pressure fluctuation factor Ybyz1 and the steam pipe pressure fluctuation factor Ybyz2 will be calculated based on the standard deviation. The standard deviation reflects the discreteness of the pressure signal, that is, the deviation of the pressure value from its average value.

[0104] The identification unit is used to determine the time point when the current corresponding abnormal unit was marked as an abnormal unit last time, marking it as the starting timestamp, and marking the current time as the ending timestamp, based on the starting timestamp and the ending timestamp, obtain the time interval, use the time interval as the comparison period, and extract the fault assessment index Ggzs at the starting timestamp from the historical data group according to the starting timestamp h , by taking the current fault assessment index Ggzs of each unit during operation as c The fault assessment index Ggzs at the start timestamp h Compare to determine whether the corresponding abnormal unit needs maintenance. The specific contents are as follows:

[0105] If the current fault assessment index Ggzs of each unit during operation c The fault assessment index Ggzs exceeding the start timestamp h When the maintenance order is issued, maintenance personnel will be arranged to carry out on-site maintenance work;

[0106] If the current fault assessment index Ggzs of each unit during operation c The fault assessment index Ggzs at the start timestamp is not exceeded h At this time, no maintenance instruction will be issued.

[0107] In this embodiment, after receiving the trend analysis instruction, the fault analysis unit uses the equipment status data in the real-time data group to eliminate the dimension difference of the data through dimensionless processing, so as to ensure that the fault characteristic parameters of different properties can be analyzed under the same standard. The current fault assessment index is calculated by the formula, and the multi-dimensional data such as the unit startup priority difference, boiler pressure fluctuation factor, steam pipeline pressure fluctuation factor and unit load fluctuation factor are combined, and different weights are assigned to each item to comprehensively evaluate the current fault risk of the unit. The fault assessment method comprehensively considers multiple core parameters that affect the operation stability of the unit, avoids the misjudgment or omission that may be caused by the judgment of a single parameter, realizes the quantitative analysis of the fault risk of the abnormal unit, and significantly improves the scientificity and accuracy of the fault assessment. The identification unit combines the current time point and historical data to dynamically track the fault change trend of the abnormal unit. Specifically: the time point when the abnormal unit was marked as abnormal last time is used as the starting timestamp, and the current time is marked as the end timestamp, the time interval is calculated, and the historical data in the time period is extracted. By comparing the current fault assessment index with the fault assessment index at the starting timestamp, the fault risk development trend of the unit is accurately judged. If the fault assessment index continues to increase and exceeds the historical value, the system will immediately issue a maintenance order to ensure that the fault problem can be handled in a timely manner. If the fault assessment index does not exceed the historical value, the system will suspend the maintenance order to avoid unnecessary maintenance operations. This dynamic fault analysis method based on time intervals can significantly reduce the waste of resources caused by excessive maintenance, while ensuring that potential faults can be handled in a timely manner before the problem worsens.

[0108] Example 6

[0109] Please refer to Figure 1 ,Specifically: the effect feedback and optimization module includes an effect feedback unit and an optimization unit;

[0110] The effect feedback unit is used to provide feedback according to the actual effect formed by the startup sequence of each unit, and obtain relevant feedback data, wherein the relevant feedback data includes the number m of normal units, the load output value of the corresponding normal units at each moment, the power generation load target value of the corresponding normal units at each moment, and the fuel consumption of the corresponding normal units at each moment; through the relevant feedback data, the balance effect between the load response capability of the power grid and the operating efficiency of the units is evaluated, and after dimensionless processing, a performance feedback index Xfzs is constructed, and the performance feedback index Xfzs is obtained by the following formula:

[0111]

[0112] Where k = 1, 2, ..., m, m represents the number of normal units, Xfzs(t) represents the performance feedback index at time t, and Fsz k(t) represents the load output value of the kth normal unit at time t, Fbz k (t) represents the target power generation load value of the kth normal unit at time t, Rxz k (t) represents the fuel consumption of the kth normal unit at time t, ω represents the weight coefficient, which is used to balance the relationship between load error and fuel consumption; |Fsz k (t)-Fbz k (t)| represents the deviation between the actual load output and the target load; Reflects fuel economy. The smaller the value, the less fuel is consumed per unit of power generation, and the better the economy.

[0113] The optimization unit is used to adjust and optimize the load of the corresponding unit by using the gradient descent method and combining the performance feedback index Xfzs to obtain the load adjustment amount Ft of the corresponding unit at each subsequent moment:

[0114]

[0115] Where Ft(t) represents the performance feedback index at time t, and η represents the learning rate, which is used to control the amplitude of adjustment; Represents the partial derivative of the performance feedback index with respect to the load output value of the unit at time t.

[0116] In this embodiment, the effect feedback unit comprehensively evaluates the balance effect between the load response capability of the power grid and the operating efficiency of the unit by real-time collection and dimensionless processing of the operating data of each normal unit (including load output value, power generation load target value, and fuel consumption). The performance feedback index Xfzs constructed by the formula comprehensively considers the deviation between the load output of the unit and the target load, as well as the fuel economy (fuel consumption per unit of power generation), and provides a scientific quantitative standard for optimizing the operation of the unit. The index can not only reflect the current operating state of the unit, but also help analyze whether the load distribution strategy is efficient, and provide an important input basis for the gradient adjustment of the optimization unit. By calculating the partial derivative of the performance feedback index with respect to the load output value of the unit, the key variables affecting the operating efficiency are accurately located, and the adjustment range is controlled by the learning rate to avoid instability caused by excessive adjustment. The load distribution is gradually optimized by the gradient descent algorithm. The system can dynamically adjust the load amount Ft of the unit at each subsequent moment to ensure that the load distribution achieves the optimal economy while meeting the power grid demand, and significantly improves the adjustment efficiency and accuracy of the load distribution. The effect feedback unit and the optimization unit form a closed-loop feedback mechanism. Through the real-time calculation of the performance feedback index Xfzs and the dynamic adjustment of the gradient descent method, it is ensured that the system can continuously optimize the load distribution strategy according to the operating effect. This closed-loop mechanism effectively avoids the limitations brought by static parameters in the traditional load distribution strategy, so that the system can flexibly respond to the dynamic changes of grid load fluctuations and unit operating status, and improve the overall operation stability and economy. The present invention realizes the scientific quantification of unit operating efficiency and grid load response capability through the construction of the performance feedback index Xfzs, and dynamically optimizes the unit load distribution based on this, significantly improving the economy and flexibility of coal-fired power generation units. The optimization strategy of the gradient descent method can quickly respond to changes in the operating status and dynamically adjust the unit load distribution, avoiding the problems of over-adjustment or lagging adjustment in traditional strategies, and enhancing the accuracy and flexibility of load distribution.

[0117] Example 7

[0118] Please refer to Figure 2 , specifically: A method for optimizing the operation of a coal-fired power generation unit based on industrial big data, comprising the following steps,

[0119] S1. Use multiple sets of monitoring instruments to monitor the relevant unit operation data and relevant equipment status data of each coal-fired power generation unit in real time, generate a real-time data set, and obtain the historical data set of each coal-fired power generation unit in the historical period based on big data technology;

[0120] S2, preprocessing the relevant data in the real-time data group, removing noise, eliminating outliers, and filling in missing values, and constructing a processed real-time data group after dimensionless processing;

[0121] S3. Analyze and judge the current startup sequence of each unit based on the processed real-time data group's relevant unit operation data and combined with the grid load demand Dfyq to obtain the unit startup priority Qyxj. According to the numerical value of the unit startup priority Qyxj, determine the startup sequence of each unit and whether there is a failure risk. If there is a failure risk, trigger the trend analysis instruction.

[0122] S4, receiving trend analysis instructions, analyzing the current fault assessment index Ggzs of each unit during operation according to the relevant equipment status data in the processed real-time data group c , and combined with the historical data group, determine whether the corresponding unit needs maintenance work;

[0123] S5. Provide feedback based on the actual effect of the startup sequence of each unit, obtain relevant feedback data, and evaluate the balance between the grid load response capability and the unit operating efficiency based on the relevant feedback data to construct a performance feedback index Xfzs. Based on the performance feedback index Xfzs, adjust and optimize the load of the corresponding unit.

[0124] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A coal-fired power generation unit operation optimization system based on industrial big data, characterized by: It includes data acquisition module, data processing module, startup priority analysis module, fault identification module and effect feedback and optimization module; The data acquisition module is used to use multiple groups of monitoring instruments to monitor the relevant unit operation data and relevant equipment status data of each coal-fired power generation unit in real time, generate a real-time data group, and obtain the historical data group of each coal-fired power generation unit in a historical period based on big data technology; The data processing module is used to perform data preprocessing on the relevant data in the real-time data group, remove noise, eliminate outliers, and fill in missing values, and after dimensionless processing, construct a processed real-time data group; The startup priority analysis module is used to analyze and determine the startup sequence of each unit based on the processed real-time data group's related unit operation data and combined with the power grid load demand Dfyq to obtain the unit startup priority Qyxj. According to the numerical value of the unit startup priority Qyxj, the startup sequence of each unit and whether there is a fault risk are determined. If there is a fault risk, the trend analysis instruction is triggered; The fault identification module is used to receive trend analysis instructions and analyze the current fault assessment index Ggzs of each unit during operation according to the processed relevant equipment status data in the real-time data group. c , and combined with the historical data group, determine whether the corresponding unit needs maintenance work; The effect feedback and optimization module is used to provide feedback based on the actual effect formed by the startup sequence of each unit, obtain relevant feedback data, and evaluate the balance effect between the grid load response capability and the unit operating efficiency based on the relevant feedback data to construct a performance feedback index Xfzs. Based on the performance feedback index Xfzs, the load of the corresponding unit is adjusted and optimized.

2. According to the coal-fired power generation unit operation optimization system based on industrial big data according to claim 1, it is characterized by: The data acquisition module includes a real-time monitoring unit and a history extraction unit; The real-time monitoring unit is used to monitor the relevant unit operation data and relevant equipment status data of each coal-fired power generation unit in real time using multiple sets of monitoring instruments, wherein the relevant unit operation data includes the unit time fuel consumption Rxz, load output value Fsz, availability coefficient Kyz of the corresponding unit at each time, unit load fluctuation factor Fbyz and maximum power generation load output value Fbz of the corresponding unit max ; The relevant equipment status data includes boiler pressure fluctuation factor Ybyz1 and steam pipeline pressure fluctuation factor Ybyz1; multiple groups of monitoring instruments include laser coal flow meter, power monitoring instrument and unit performance test instrument; the real-time data group includes relevant unit operation data and relevant equipment status data; The history extraction unit is used to obtain relevant unit operation data and relevant equipment status data of each coal-fired power generation unit in the industrial park within a historical period using big data technology. The historical data group includes relevant unit operation data and relevant equipment status data within the historical period.

3. The coal-fired power generation unit operation optimization system based on industrial big data according to claim 2 is characterized in that: The data processing module includes a preprocessing unit and a normalization unit; The preprocessing unit is used to preprocess the relevant data in the real-time data group, and the preprocessing includes removing noise, filling missing values, removing outliers and data smoothing operations, wherein the missing value filling methods include mean filling, median filling, interpolation filling and regression filling; The normalization unit is used to eliminate unit differences of relevant data in the preprocessed real-time data group by using dimensionless processing technology, so that the range of relevant data in the preprocessed real-time data group falls within [0, 1].

4. The coal-fired power generation unit operation optimization system based on industrial big data according to claim 3 is characterized by: The startup priority analysis module includes an operation status analysis unit, a priority unit and a preliminary judgment unit; The operation status analysis unit is used to extract the unit time fuel consumption Rxz and the load output value Fsz of each unit from the processed real-time data group of the relevant unit operation data, and calculate the economic index Jzb of each unit through the unit time fuel consumption Rxz and the load output value Fsz of each unit. The economic index Jzb of each unit is obtained specifically according to the following formula: According to the economic index Jzb of each unit and the load demand of the power grid Dfyq, the target value of the power generation load Fbz of each unit in different monitoring periods is calculated. Specifically, the target value of the power generation load Fbz of each unit in different monitoring periods is obtained according to the following formula: In the formula, Fbz(t) represents the target power generation load value of the corresponding unit at time t, Dfyq(t) represents the grid load demand at time t, and Jzb -1 represents the economic weight of the corresponding unit; n represents the number of units, i = 1, 2, 3, ..., n.

5. The coal-fired power generation unit operation optimization system based on industrial big data according to claim 4 is characterized in that: The priority unit is used to analyze and determine the current startup sequence of each unit based on the processed real-time data group related unit operation data and the current grid load demand to obtain the unit startup priority Qyxj. The unit startup priority Qyxj is specifically obtained by the following formula: Where Qyxj(t) represents the unit startup priority of the corresponding unit at time t, Kyz(t) represents the availability coefficient of the corresponding unit at time t, and Fbz max It is represented by the maximum power generation load output value of the corresponding unit, and ∈ is represented by the correction constant.

6. The coal-fired power generation unit operation optimization system based on industrial big data according to claim 5 is characterized by: The preliminary judgment unit is used to obtain the unit startup priority Qyxj of each unit according to the method of obtaining the unit startup priority Qyxj in the priority unit, and sort them to generate a sequence group; determine the unit startup priority Qyxj of the corresponding unit in the historical period according to the historical data group, and obtain the average unit startup priority Qyxj of the corresponding unit in combination with the statistical averaging algorithm avg , by comparing the unit startup priority Qyxj of the current unit with the average unit startup priority Qyxj of the corresponding unit avg Compare to determine the startup sequence of the corresponding units and whether there is a risk of failure. The specific preliminary judgment is as follows: If the unit startup priority Qyxj of the current unit ≥ the average unit startup priority Qyxj of the corresponding unit avg When the corresponding unit is initially determined to have no failure risk, the trend analysis command is not triggered externally, the corresponding unit is retained in the sequence group, and the corresponding unit is marked as a normal unit; If the unit startup priority Qyxj of the current unit is less than the average unit startup priority Qyxj of the corresponding unit avg When the corresponding unit is initially determined to have a failure risk, a trend analysis instruction is triggered externally to remove the corresponding unit from the sequence group and mark the corresponding unit as an abnormal unit; According to the preliminary judgment, the normal units are counted and re-sorted to generate an optimized sequence group. Based on the optimized sequence group, the startup order of each normal unit is determined.

7. The coal-fired power generation unit operation optimization system based on industrial big data according to claim 3 is characterized by: The fault identification module includes a fault analysis unit and an identification unit; The fault analysis unit is used to analyze the current fault assessment index Ggzs of each abnormal unit during operation according to the processed relevant equipment status data in the real-time data group after receiving the trend analysis instruction and after dimensionless processing. c , the current fault assessment index Ggzs c Obtained by the following formula: In the formula, ΔQyxj represents the unit startup priority difference, Ybyz1 represents the boiler pressure fluctuation factor, Ybyz2 represents the steam pipeline pressure fluctuation factor, Fbyz represents the unit load fluctuation factor, α, β, γ and All represent weight values, where 0<α<1, 0<β<1, 0<γ<1, α, β, γ and The specific value is set by the user according to the situation; The identification unit is used to determine the time point when the current corresponding abnormal unit was marked as an abnormal unit last time, marking it as the starting timestamp, and marking the current time as the ending timestamp, based on the starting timestamp and the ending timestamp, obtain the time interval, use the time interval as the comparison period, and extract the fault assessment index Ggzs at the starting timestamp from the historical data group according to the starting timestamp h , by taking the current fault assessment index Ggzs of each unit during operation as c The fault assessment index Ggzs at the start timestamp h Compare to determine whether the corresponding abnormal unit needs maintenance. The specific contents are as follows: If the current fault assessment index Ggzs of each unit during operation c The fault assessment index Ggzs exceeding the start timestamp h When the maintenance order is issued, maintenance personnel will be arranged to carry out on-site maintenance work; If the current fault assessment index Ggzs of each unit during operation c The fault assessment index Ggzs at the start timestamp is not exceeded h At this time, no maintenance instruction will be issued.

8. The coal-fired power generation unit operation optimization system based on industrial big data according to claim 1 is characterized by: The effect feedback and optimization module includes an effect feedback unit and an optimization unit; The effect feedback unit is used to provide feedback according to the actual effect formed by the startup sequence of each unit, and obtain relevant feedback data, wherein the relevant feedback data includes the number m of normal units, the load output value of the corresponding normal units at each moment, the power generation load target value of the corresponding normal units at each moment, and the fuel consumption of the corresponding normal units at each moment; through the relevant feedback data, the balance effect between the load response capability of the power grid and the operating efficiency of the units is evaluated, and after dimensionless processing, a performance feedback index Xfzs is constructed, and the performance feedback index Xfzs is obtained by the following formula: Where k = 1, 2, ..., m, m represents the number of normal units, Xfzs(t) represents the performance feedback index at time t, and Fsz k (t) represents the load output value of the kth normal unit at time t, Fbz k (t) represents the target power generation load value of the kth normal unit at time t, Rxz k (t) represents the fuel consumption of the kth normal unit at time t, and ω represents the weight coefficient.

9. The coal-fired power generation unit operation optimization system based on industrial big data according to claim 8 is characterized by: The optimization unit is used to optimize the load adjustment of the corresponding unit by using the gradient descent method and combining the performance feedback index Xfzs to obtain the load adjustment amount Ft of the corresponding unit at each subsequent moment: Where Ft(t) represents the performance feedback index at time t, and η represents the learning rate; Represents the partial derivative of the performance feedback index with respect to the load output value of the unit at time t.

10. A method for optimizing the operation of a coal-fired power generation unit based on industrial big data, used to implement the system for optimizing the operation of a coal-fired power generation unit based on industrial big data as described in any one of claims 1 to 9, characterized in that: The following steps are included: S1. Use multiple sets of monitoring instruments to monitor the relevant unit operation data and relevant equipment status data of each coal-fired power generation unit in real time, generate a real-time data set, and obtain the historical data set of each coal-fired power generation unit in the historical period based on big data technology; S2, preprocessing the relevant data in the real-time data group, removing noise, eliminating outliers, and filling in missing values, and constructing a processed real-time data group after dimensionless processing; S3. Analyze and judge the current startup sequence of each unit based on the processed real-time data group's relevant unit operation data and combined with the grid load demand Dfyq to obtain the unit startup priority Qyxj. According to the numerical value of the unit startup priority Qyxj, determine the startup sequence of each unit and whether there is a failure risk. If there is a failure risk, trigger the trend analysis instruction. S4, receiving trend analysis instructions, analyzing the current fault assessment index Ggzs of each unit during operation according to the relevant equipment status data in the processed real-time data group c , and combined with the historical data group, determine whether the corresponding unit needs maintenance work; S5. Provide feedback based on the actual effect of the startup sequence of each unit, obtain relevant feedback data, and evaluate the balance between the grid load response capability and the unit operating efficiency based on the relevant feedback data to construct a performance feedback index Xfzs. Based on the performance feedback index Xfzs, adjust and optimize the load of the corresponding unit.

Images