Power operation and maintenance management method and system based on big data

Abstract

The invention discloses an electric power operation and maintenance management method and system based on big data, and the method comprises the following steps: collecting electric power data, and carrying out the data preprocessing; calculating the power generation unit efficiency ratio according to the preprocessed data; predicting the power demand, and calculating a power dispatching scheme set by using an optimization algorithm based on the power generation unit efficiency ratio and the power demand prediction; calculating a power generation machine fault probability, and calculating an implementation probability of each scheme in the power dispatching scheme set according to the machine fault probability; and selecting an optimal power dispatching scheme, and performing operation and maintenance effect evaluation and feedback circulation. The electric power data is analyzed based on the big data, the operation cost is minimized, the electric power requirement is met, and the economical efficiency and flexibility of the electric power system are improved.

Description

Technical Field

[0001] The present invention relates to the field of data processing, and in particular to a power operation and maintenance management method and system based on big data. Background Art

[0002] As power systems continue to expand and technology advances, power operations and maintenance (O&M) management faces increasingly complex challenges. Traditional O&M management methods rely primarily on manual experience and regular inspections. Modern power systems involve multiple subsystems and a large number of devices, making traditional manual management methods difficult to cope with this complexity. This approach is not only inefficient but also struggles to cope with sudden outages and the demands of large-scale data analysis.

[0003] The development of big data technology has brought new opportunities to power operations and maintenance management. Big data-based power operations and maintenance management methods improve power system efficiency, reduce costs, and enhance reliability and service quality through efficient data collection, storage, processing, and analysis. Faced with increasingly complex power systems and massive amounts of data, the application of big data technology has become an inevitable trend and is of great significance for promoting the sustainable development of the power industry. Summary of the Invention

[0004] The present invention provides a power operation and maintenance management method based on big data, comprising:

[0005] S10, collecting power data and performing data preprocessing;

[0006] S20, calculating the power generation unit efficiency ratio based on the preprocessed data;

[0007] S30, predicting power demand, and calculating a set of power dispatching solutions using an optimization algorithm based on the power generation unit efficiency ratio and the power demand forecast;

[0008] S40, calculating the probability of a generator failure, and calculating the probability of implementation of each plan in the power dispatch plan set based on the machine failure probability;

[0009] S50. Select the optimal power dispatching plan and conduct operation and maintenance effect evaluation and feedback loop.

[0010] In the power operation and maintenance management method based on big data as described above, the collected power data includes power system data, smart meter data, power generation equipment data and external meteorological data.

[0011] In the power operation and maintenance management method based on big data as described above, data preprocessing of the collected data includes aligning timestamps of data from different sources, detecting and processing outliers, and using linear interpolation to fill missing values in the data.

[0012] As described above, the power operation and maintenance management method based on big data calculates the efficiency ratio of the power generation unit to the actual power generation efficiency of the required power generation unit and the theoretical maximum power generation efficiency of the power generation unit, sorts and analyzes the power generation units according to the efficiency ratio, and uses the percentile method to identify the power generation units with lower efficiency.

[0013] As described above, in the power operation and maintenance management method based on big data, the power demand forecasting needs to be based on the average power consumption of users at the current time, taking into account the impact of power demand at past time points on the current demand and the impact of past error terms on the current demand.

[0014] As described above, the power operation and maintenance management method based on big data, in which the optimization algorithm is used to solve the scheduling plan set, is to use the optimization algorithm to minimize the cost function. The specific steps include randomly generating a set of initial solutions to form an initial population; using the fitness function to evaluate whether the individuals in the population meet the constraints; and the individuals that meet the constraints form a scheduling plan set.

[0015] As described above, in the power operation and maintenance management method based on big data, the calculation of the realization probability of each plan in the power dispatching plan set needs to consider the probability of machine failure, whether the combination of power generation units in the dispatching plan can be actually realized, the number of equipment maintenance times, and the age of the equipment.

[0016] The present invention also provides a carbon emission data quality early warning system based on big data, including:

[0017] Acquisition module: used to collect power data and perform data preprocessing;

[0018] Efficiency calculation module: used to calculate the efficiency ratio of the power generation unit based on the preprocessed data;

[0019] Power dispatch plan set module: used to predict power demand, based on the power generation unit efficiency ratio and power demand forecast, and use optimization algorithm to calculate the power dispatch plan set;

[0020] Scheme probability calculation module: used to calculate the probability of generator failure and calculate the realization probability of each scheme in the power dispatch scheme set based on the machine failure probability;

[0021] Solution effect module: select the optimal power dispatch solution, conduct operation and maintenance effect evaluation and feedback loop.

[0022] The beneficial effects achieved by the present invention are as follows: the present invention analyzes power data based on big data, minimizes operating costs while meeting power demand, and improves the economy and flexibility of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments described in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0024] Figure 1 This is a flow chart of the power operation and maintenance management method based on big data provided in Example 1 of this application.

[0025] Figure 2 This is a schematic diagram of the big data-based power operation and maintenance management system provided in Example 2 of the present application. DETAILED DESCRIPTION

[0026] The following is a clear and complete description of the technical solutions in the embodiments of the present invention, in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0027] Example 1

[0028] like Figure 1 As shown, the first embodiment of the present application provides a power operation and maintenance management method based on big data, including:

[0029] S10: Collect power data and perform data preprocessing.

[0030] Obtain various physical parameters of the power system, including but not limited to the power generation units generating electricity Energy consumed by the power generation unit Power generation facility machine data. This data is obtained from user-side smart meters. The data format of smart meters typically includes a timestamp, power consumption, and user ID. The timestamp is parsed to obtain the power consumption time t. This data allows for a more precise understanding of user electricity usage habits, thereby optimizing power supply strategies and service quality.

[0031] Align the timestamps of data from different sources, including power system data, smart meter data, and external meteorological data. Detect and process outliers in these data. Outliers are observations that are significantly different from other data points. The formula for detecting outliers is: Where X is the observed value, μ is the mean, and σ is the standard deviation. When |Z| > 3, the value is considered an outlier and can be deleted or replaced with the mean. Outlier handling can significantly improve data quality and avoid misjudgments in subsequent analyses.

[0032] For missing values in the data, linear interpolation is used to fill them and estimate the missing values through known adjacent observations. The specific formula is: where Y missing Represents the estimated value of the missing value, that is, at time t missing The estimated value of the missing data at is the known data point Y i Add the value from Y i to Y missing The increment of Y i and Y j Represents two known adjacent observations, t i Represents a known data point Y i The corresponding time point, t j Represents a known data point Y j The corresponding time point, t missing Indicates the time point corresponding to the missing value, It is the slope between two known data points, indicating the rate of change of the data during this period. Represents a known data point Y i to the missing data point Y missing Interpolation can effectively fill in the data gaps in a short period of time and maintain data continuity.

[0033] S20. Calculate the power generation unit efficiency ratio based on the preprocessed data.

[0034] Extract the data of each power generation unit from the preprocessed data and calculate the efficiency ratio E of each power generation unit. The formula is: in, Indicates the electrical energy generated by the power generation unit, Indicates the energy consumed by the power generation unit, Indicates the actual power generation efficiency of the power generation unit, E max This represents the theoretical maximum efficiency of a power generation unit, which can be obtained from the equipment nameplate or design parameters. E is a value between 0 and 1. Values closer to 1 indicate a more efficient power generation unit, while values lower than 1 indicate potential performance issues.

[0035] According to the calculated efficiency ratio, the power generation units are sorted and analyzed to identify the power generation units with lower efficiency. Using the percentile method, calculate the ath percentile B a The ath percentile means that a% of the data points in the data set are less than or equal to this value. Arrange all the efficiency ratios E in ascending order and use the formula Calculates the ath percentile position, where n is the number of data points in the dataset. If y is an integer, then the ath percentile is the value of the yth data point. If y is not an integer, then the ath percentile is the value of the data points and The linear interpolation of data points is The specific formula is yh represents the index difference between the interpolated point y and the known point h, and kh represents the index difference between the known point k and the known point h. Is a scaling factor that represents the relative position of the interpolated point y between the known points h and k. k is the efficiency ratio of point k, E h is the efficiency ratio of point h.

[0036] By analyzing the distribution of efficiency ratios, we can identify inefficient or abnormal power generation units and conduct further inspection and maintenance. b , you need to focus on its operating status, find possible problems and take corresponding measures.

[0037] S30. Predicting power demand, and using an optimization algorithm to calculate a set of power dispatching solutions based on the power generation unit efficiency ratio and power demand forecast.

[0038] S31. Forecast future electricity demand.

[0039] Power demand is the basis of power dispatching. Time series analysis method is used to predict the future power demand of users. The calculation formula is P forecast =Q+φ1P t-1 +φ2P t-2 +...+φ p P t-p +θ1ε t-1 +θ2ε t-2 +...+θ q ε t-q , where P forecast represents the predicted value of the user's electricity demand at a certain time in the future, Q represents the average value of the user's electricity consumption at the current time t, φ p represents the autoregressive coefficient, which is used to measure the impact of electricity demand at past time points on current demand. φ1 represents the impact of electricity demand at time t-1 on current demand. P t-1 represents the power demand value at time t-1, P t-p θ represents the power demand value at time tp, p represents the order of the autoregressive term, and is the number of past power demand values. q represents the moving average coefficient, which is used to measure the impact of past error terms on current demand, θ1 represents the impact of the error term at time t-1 on current demand, and ε t-1 represents the prediction error at time t-1, ε t-iThe term t represents the forecast error at time ti, which is the difference between the actual value and the forecast value. q represents the order of the moving average term, which is the number of past error terms. By predicting users' future electricity demand, we can make scheduling preparations in advance and ensure the stability and reliability of power supply.

[0040] S32. Construct a cost function.

[0041] Constructing the cost function C can quantify the cost of different scheduling schemes and provide a basis for optimizing scheduling. Factors such as power generation cost, transmission loss and environmental impact need to be considered during the construction process. The calculation formula of the cost function is where c i , d i , e i and f i is the cost coefficient, is the power generation of the i-th power generation unit, and r represents the number of power generation units. Indicates linearly increasing costs such as fuel costs and labor costs, Represents secondary costs such as equipment wear and maintenance costs, The square root cost of a device that is more efficient at low loads but less efficient at high loads. It represents the logarithmic cost of certain environmental measures that are more effective at low power generation but less effective at high power generation. represents the cost of transmission losses per unit of electricity generated, Represents the cost of environmental impact per unit of electricity generated.

[0042] S33. Use an optimization algorithm to solve a set of scheduling solutions.

[0043] Using an optimization algorithm to solve a dispatch solution involves minimizing the cost function C. This optimization process requires consideration of multiple constraints: power generation cannot exceed the theoretical maximum, and power demand must be met. The optimization algorithm iteratively searches to find the optimal solution that satisfies all constraints, thereby optimizing power dispatch. This minimizes operating costs while meeting power demand, improving the economics and flexibility of the power system.

[0044] S331. Initialize the population.

[0045] A set of initial solutions is randomly generated, and the number of initial solutions is as large as possible to form an initial population. The population consists of individuals, each of which consists of multiple power generation units, and each individual is a potential solution. Let N be the population size, and the initial population can be expressed as in, It is the Nth individual, and the composition of the power generation unit of this individual is random.

[0046] S332. Evaluate fitness.

[0047] Check whether each individual satisfies all constraints. Only select individuals that satisfy all constraints.

[0048] For each individual's fitness function:

[0049] fitness(x) represents the fitness value of individual x, which reflects the degree of performance of the individual in the optimization problem. In the minimization problem, the lower the fitness value, the better the individual. m represents the number of power generation units used by the individual. represents the cost function of the i-th power generation unit, E i Represents the efficiency ratio of the i-th power generation unit, α1 represents the penalty factor of the equality constraint, which is a non-negative constant used to control the degree of penalty when the equality constraint is violated. represents the electrical energy provided by a single power generation unit in individual x, represents the total electrical energy provided by individual x, It represents the total electricity required by the users, and u is the number of users. Represents the deviation of power demand, the smaller the deviation, the better. β1 represents the penalty factor of the inequality constraint, which is a non-negative constant used to control the degree of penalty when the inequality constraint is violated. represents the total penalty of the inequality constraint. For each power generation unit i, if but Indicates that no inequality constraints are violated. M i is the theoretical maximum power generation of the i-th power generation unit.

[0050] The individuals that meet the conditions are formed into a new population F1. If there are no individuals that meet the conditions, two individuals are randomly selected for crossover to generate new offspring individuals. ψ Mutate individuals to increase population diversity. Add newly generated offspring individuals to the population, replacing some individuals with lower fitness, to form a new generation of population. Use the fitness function to filter until qualified individuals are found, forming a set of scheduling solutions.

[0051] In the set of power dispatching schemes, there are many different combinations of power generation units that can meet power demand and achieve the same minimum cost, but the specific allocation of power generation units is different.

[0052] S40. Calculate the probability of a generator failure, and calculate the probability of achieving each plan in the power dispatching plan set based on the machine failure probability.

[0053] By analyzing the historical operation data and current status information of the equipment, the future failure probability of the equipment is predicted to evaluate the health status of the equipment. The historical operation data of the equipment, including the operation time, is extracted from the preprocessed data. t , number of failures d f , maintenance records, equipment age, etc. At the same time, the current equipment status information is collected, including temperature T, vibration V, current I, etc. This data can be obtained through sensors and historical records.

[0054] Predicting the probability of equipment failure is a key step in preventive maintenance. The calculation formula is Where μ0 represents the intercept term, the basic failure probability when all other variables are 0, μ1 represents the effect of temperature T on the failure probability, μ2 represents the effect of voltage V on the failure probability, μ3 represents the effect of current I on the failure probability, and μ4 represents the number of failures d f Impact on the failure probability, μ5 represents the running time d t Impact on the probability of failure.

[0055] The realization probability of the i-th plan in the power dispatch plan set is calculated based on the machine failure probability. The specific formula is D actual_i =d realize_i ×(1-R)×(1+θ×age i )×(1+τ×record i ). Among them, d realize_i Indicates that the combination of power generation units in the i-th scheduling plan can actually be realized, 1-R represents the probability of normal operation of the machine, θ represents the weight of the equipment age, age i Indicates that the scheduling scheme used is different and the age of the equipment used is different. τ represents the weight of the equipment maintenance times, record i This means that different equipment has different maintenance times. The more times a device has been repaired, the lower the probability of the solution being implemented.

[0056] S50. Select the optimal power dispatching plan and conduct operation and maintenance effect evaluation and feedback loop.

[0057] According to the realization probability of each plan, the power dispatch plan with the highest realization probability is selected. The cost saving rate after applying this dispatch plan is R cost =(R before -R after ) / R before , where R before is the cost before optimization, R after To obtain the optimized cost, the cost coefficient in the optimization cost function can be adjusted according to the cost saving rate.

[0058] After implementing this solution, the effectiveness of power operation and maintenance management should be regularly evaluated. By comparing actual operating data with the solution's predicted values, algorithm parameters should be adjusted and the operation and maintenance management strategy optimized, forming a closed loop of continuous improvement. Power generation units with excessively low efficiency or aging should be replaced, and power generation equipment should be regularly inspected and maintained to avoid plan errors caused by equipment problems. User feedback and new data should be collected to understand their needs and expectations, allowing timely adjustments to the operation and maintenance strategy to improve user satisfaction and service quality. Based on this information, an operation and maintenance report should be generated, including an evaluation of operation and maintenance results, problem analysis, and improvement suggestions.

[0059] Example 2

[0060] like Figure 2 As shown, the second embodiment of the present application provides a power operation and maintenance management system based on big data, including:

[0061] Acquisition module: used to collect power data and perform data preprocessing.

[0062] Obtain various physical parameters of the power system, including but not limited to the power generation units generating electricity Energy consumed by the power generation unit Power generation facility machine data. This data is obtained from user-side smart meters. The data format of smart meters typically includes a timestamp, power consumption, and user ID. The timestamp is parsed to obtain the power consumption time t. This data allows for a more precise understanding of user electricity usage habits, thereby optimizing power supply strategies and service quality.

[0063] Align the timestamps of data from different sources, including power system data, smart meter data, and external meteorological data. Detect and process outliers in these data. Outliers are observations that are significantly different from other data points. The formula for detecting outliers is: Where X is the observed value, μ is the mean, and σ is the standard deviation. When |Z| > 3, the value is considered an outlier and can be deleted or replaced with the mean. Outlier handling can significantly improve data quality and avoid misjudgments in subsequent analyses.

[0064] For missing values in the data, linear interpolation is used to fill them and estimate the missing values through known adjacent observations. The specific formula is: where Y missing Represents the estimated value of the missing value, that is, at time t missing The estimated value of the missing data at is the known data point Y i Add the value from Y i to Y missing The increment of Y i and Y j Represents two known adjacent observations, t i Represents a known data point Yi The corresponding time point, t j Represents a known data point Y j The corresponding time point, t missing Indicates the time point corresponding to the missing value, It is the slope between two known data points, indicating the rate of change of the data during this period. Represents a known data point Y i to the missing data point Y missing Interpolation can effectively fill in the data gaps in a short period of time and maintain data continuity.

[0065] Efficiency calculation module: used to calculate the efficiency ratio of the power generation unit based on the preprocessed data.

[0066] Extract the data of each power generation unit from the preprocessed data and calculate the efficiency ratio E of each power generation unit. The formula is: in, Indicates the electrical energy generated by the power generation unit, Indicates the energy consumed by the power generation unit, Indicates the actual power generation efficiency of the power generation unit, E max This represents the theoretical maximum efficiency of a power generation unit, which can be obtained from the equipment nameplate or design parameters. E is a value between 0 and 1. Values closer to 1 indicate a more efficient power generation unit, while values lower than 1 indicate potential performance issues.

[0067] According to the calculated efficiency ratio, the power generation units are sorted and analyzed to identify the power generation units with lower efficiency. Using the percentile method, calculate the ath percentile B a The ath percentile means that a% of the data points in the data set are less than or equal to this value. Arrange all the efficiency ratios E in ascending order and use the formula Calculates the ath percentile position, where n is the number of data points in the dataset. If y is an integer, then the ath percentile is the value of the yth data point. If y is not an integer, then the ath percentile is the value of the data points and Linear interpolation of data points, let The specific formula is yh represents the index difference between the interpolated point y and the known point h, and kh represents the index difference between the known point k and the known point h. Is a scaling factor that represents the relative position of the interpolated point y between the known points h and k. k is the efficiency ratio of point k, E h is the efficiency ratio of point h.

[0068] By analyzing the distribution of efficiency ratios, we can identify inefficient or abnormal power generation units and conduct further inspection and maintenance. b , you need to focus on its operating status, find possible problems and take corresponding measures.

[0069] Power dispatching scheme collection module: including power demand forecasting submodule, cost function submodule, and scheme collection submodule.

[0070] Power demand forecasting submodule: used to forecast future power demand.

[0071] Power demand is the basis of power dispatching. Time series analysis method is used to predict the future power demand of users. The calculation formula is P forecast =Q+φ1P t-1 +φ2P t-2 +...+φ p P t-p +θ1ε t-1 +θ2ε t-2 +...+θ q ε t-q , where P forecast represents the predicted value of the user's electricity demand at a certain time in the future, Q represents the average value of the user's electricity consumption at the current time t, φ p represents the autoregressive coefficient, which is used to measure the impact of electricity demand at past time points on current demand. φ1 represents the impact of electricity demand at time t-1 on current demand. P t-1 represents the power demand value at time t-1, P t-p θ represents the power demand value at time tp, p represents the order of the autoregressive term, and is the number of past power demand values. q represents the moving average coefficient, which is used to measure the impact of past error terms on current demand, θ1 represents the impact of the error term at time t-1 on current demand, and ε t-1 represents the prediction error at time t-1, ε t-i The term t represents the forecast error at time ti, which is the difference between the actual value and the forecast value. q represents the order of the moving average term, which is the number of past error terms. By predicting users' future electricity demand, we can make scheduling preparations in advance and ensure the stability and reliability of power supply.

[0072] Cost function submodule: used to construct the cost function.

[0073] Constructing the cost function C can quantify the cost of different scheduling schemes and provide a basis for optimizing scheduling. Factors such as power generation cost, transmission loss and environmental impact need to be considered during the construction process. The calculation formula of the cost function is where c i , di , e i and f i is the cost coefficient, is the power generation of the i-th power generation unit, and r represents the number of power generation units. Indicates linearly increasing costs such as fuel costs and labor costs, Represents secondary costs such as equipment wear and maintenance costs, The square root cost of a device that is more efficient at low loads but less efficient at high loads. It represents the logarithmic cost of certain environmental measures that are more effective at low power generation but less effective at high power generation. represents the cost of transmission losses per unit of electricity generated, Represents the cost of environmental impact per unit of electricity generated.

[0074] Scheme collection submodule: includes initialization submodule, fitness submodule, and collection submodule.

[0075] Initialization submodule: used to initialize the population.

[0076] A set of initial solutions is randomly generated, and the number of initial solutions is as large as possible to form an initial population. The population consists of individuals, each of which consists of multiple power generation units, and each individual is a potential solution. Let N be the population size, and the initial population can be expressed as in, It is the Nth individual, and the composition of the power generation unit of this individual is random.

[0077] Fitness submodule: used to check whether each individual meets all constraints.

[0078] Using an optimization algorithm to solve a dispatch solution involves minimizing the cost function C. This optimization process requires consideration of multiple constraints: power generation cannot exceed the theoretical maximum, and power demand must be met. The optimization algorithm iteratively searches to find the optimal solution that satisfies all constraints, thereby optimizing power dispatch. This minimizes operating costs while meeting power demand, improving the economics and flexibility of the power system.

[0079] Check whether each individual satisfies all constraints. Only select individuals that satisfy all constraints.

[0080] For each individual's fitness function:

[0081] fitness(x) represents the fitness value of individual x, which reflects the degree of performance of the individual in the optimization problem. In the minimization problem, the lower the fitness value, the better the individual. m represents the number of power generation units used by the individual. represents the cost function of the i-th power generation unit, E i Represents the efficiency ratio of the i-th power generation unit, α1 represents the penalty factor of the equality constraint, which is a non-negative constant used to control the degree of penalty when the equality constraint is violated. represents the electrical energy provided by a single power generation unit in individual x, represents the total electrical energy provided by individual x, It represents the total electricity required by the users, and u is the number of users. Represents the deviation of power demand, the smaller the deviation, the better. β1 represents the penalty factor of the inequality constraint, which is a non-negative constant used to control the degree of penalty when the inequality constraint is violated. represents the total penalty of the inequality constraint. For each power generation unit i, if but Indicates that no inequality constraints are violated. M i is the theoretical maximum power generation of the i-th power generation unit.

[0082] Collection submodule: used to form a scheduling plan collection.

[0083] The individuals that meet the conditions are formed into a new population F1. If there are no individuals that meet the conditions, two individuals are randomly selected for crossover to generate new offspring individuals. ψ Mutate individuals to increase population diversity. Add newly generated offspring individuals to the population, replacing some individuals with lower fitness, to form a new generation of population. Use the fitness function to filter until qualified individuals are found, forming a set of scheduling solutions.

[0084] In the set of power dispatching schemes, there are many different combinations of power generation units that can meet power demand and achieve the same minimum cost, but the specific allocation of power generation units is different.

[0085] Solution probability calculation module: includes failure probability submodule and solution implementation probability submodule.

[0086] Failure probability submodule: used to calculate the failure probability of the equipment.

[0087] By analyzing the historical operation data and current status information of the equipment, the future failure probability of the equipment is predicted to evaluate the health status of the equipment. The historical operation data of the equipment, including the operation time, is extracted from the preprocessed data. t , number of failures d f , maintenance records, equipment age, etc. At the same time, the current equipment status information is collected, including temperature T, vibration V, current I, etc. This data can be obtained through sensors and historical records.

[0088] Predicting the probability of equipment failure is a key step in preventive maintenance. The calculation formula is Where μ0 represents the intercept term, the basic failure probability when all other variables are 0, μ1 represents the effect of temperature T on the failure probability, μ2 represents the effect of voltage V on the failure probability, μ3 represents the effect of current I on the failure probability, and μ4 represents the number of failures d f Impact on the failure probability, μ5 represents the running time d t Impact on the probability of failure.

[0089] Scheme realization probability submodule: used to calculate the realization probability of the scheme in the power dispatch scheme set based on the machine failure probability.

[0090] The realization probability of the i-th plan in the power dispatch plan set is calculated based on the machine failure probability. The specific formula is D actual_i =d realize_i ×(1-R)×(1+θ×age i )×(1+τ×record i ). Among them, d realize_i Indicates that the combination of power generation units in the i-th scheduling plan can actually be realized, 1-R represents the probability of normal operation of the machine, θ represents the weight of the equipment age, age i Indicates that the scheduling scheme used is different and the age of the equipment used is different. τ represents the weight of the equipment maintenance times, record i This means that different equipment has different maintenance times. The more times a device has been repaired, the lower the probability of the solution being implemented.

[0091] Solution effect module: select the optimal power dispatch solution, conduct operation and maintenance effect evaluation and feedback loop.

[0092] According to the realization probability of each plan, the power dispatch plan with the highest realization probability is selected. The cost saving rate after applying this dispatch plan is R cost =(R before -R after ) / R before , where R before is the cost before optimization, R after To obtain the optimized cost, the cost coefficient in the optimization cost function can be adjusted according to the cost saving rate.

[0093] After implementing this solution, the effectiveness of power operation and maintenance management should be regularly evaluated. By comparing actual operating data with the solution's predicted values, algorithm parameters should be adjusted and the operation and maintenance management strategy optimized, forming a closed loop of continuous improvement. Power generation units with excessively low efficiency or aging should be replaced, and power generation equipment should be regularly inspected and maintained to avoid plan errors caused by equipment problems. User feedback and new data should be collected to understand their needs and expectations, allowing timely adjustments to the operation and maintenance strategy to improve user satisfaction and service quality. Based on this information, an operation and maintenance report should be generated, including an evaluation of operation and maintenance results, problem analysis, and improvement suggestions.

[0094] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent replacements, improvements, etc. made on the basis of the technical solution of the present invention should be included in the scope of protection of the present invention.

Claims

1. The power operation and maintenance management method based on big data is characterized by: include: S10, collecting power data and performing data preprocessing; S20, calculating the power generation unit efficiency ratio based on the preprocessed data; S30, predicting power demand, and calculating a set of power dispatching solutions using an optimization algorithm based on the power generation unit efficiency ratio and the power demand forecast; S40, calculating the probability of a generator failure, and calculating the probability of implementation of each plan in the power dispatch plan set based on the machine failure probability; S50. Select the optimal power dispatching plan and conduct operation and maintenance effect evaluation and feedback loop.

2. The power operation and maintenance management method based on big data according to claim 1, characterized in that: The collected power data includes power system data, smart meter data, power generation equipment data and external meteorological data.

3. The power operation and maintenance management method based on big data according to claim 1, characterized in that: Data preprocessing of the collected data includes aligning the timestamps of data from different sources, detecting and processing outliers, and using linear interpolation to fill in missing values in the data.

4. The power operation and maintenance management method based on big data according to claim 1, characterized in that: Calculate the efficiency ratio of the power generation unit by comparing the actual power generation efficiency of the required power generation unit to the theoretical maximum power generation efficiency of the power generation unit. Sort and analyze the power generation units according to the efficiency ratio, and use the percentile method to identify the power generation units with lower efficiency.

5. The power operation and maintenance management method based on big data according to claim 1, characterized in that: The electricity demand forecast needs to consider the impact of electricity demand at past time points on current demand and the impact of past error terms on current demand based on the average electricity consumption of users at the current time.

6. The power operation and maintenance management method based on big data according to claim 1, characterized in that: Using an optimization algorithm to solve a set of scheduling solutions means using an optimization algorithm to minimize the cost function. The specific steps include randomly generating a set of initial solutions to form an initial population; using a fitness function to evaluate whether individuals in the population meet the constraints; and individuals that meet the constraints form a set of scheduling solutions.

7. The power operation and maintenance management method based on big data according to claim 1, characterized in that: Calculating the realization probability of each plan in the power dispatch plan set requires considering the probability of machine failure, whether the combination of power generation units in the dispatch plan can be actually realized, the number of equipment maintenance times, and the age of the equipment.

8. The power operation and maintenance management system based on big data is characterized by: include: Acquisition module: used to collect power data and perform data preprocessing; Efficiency calculation module: used to calculate the efficiency ratio of the power generation unit based on the preprocessed data; Power dispatch plan set module: used to predict power demand, based on the power generation unit efficiency ratio and power demand forecast, and use optimization algorithm to calculate the power dispatch plan set; Scheme probability calculation module: used to calculate the probability of generator failure and calculate the realization probability of each scheme in the power dispatch scheme set based on the machine failure probability; Solution effect module: select the optimal power dispatch solution, conduct operation and maintenance effect evaluation and feedback loop.

Images