Active-standby combined power grid dispatch optimization system based on cloud computing

Abstract

The present invention relates to the field of power grid dispatching technology, and specifically, to a main-standby combined power grid dispatching optimization system based on cloud computing. It includes: a data acquisition unit collects electrical data of power grid operation from equipment in the power grid, and uploads the electrical data to the cloud through wireless communication; a data storage and analysis unit stores the electrical data in the cloud, and uses a regression model to analyze the electrical data; a dispatching decision support unit uses an optimal power flow algorithm based on the electrical data in the data storage and analysis unit to formulate a power generation plan and load distribution strategy; a fault detection and recovery unit detects abnormal conditions in the power grid through real-time monitoring, quickly locates the fault point, and initiates an emergency plan to isolate and restore the fault. The present invention is designed to upload real-time electrical data to the cloud and process it using a linear regression model, so that the system can more accurately predict future load demand.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid dispatching, and in particular to a cloud computing-based active-standby combined power grid dispatching optimization system. Background Art

[0002] Traditional power grid dispatching systems usually rely on local data centers and limited computing resources. They have insufficient processing capabilities when faced with massive amounts of power grid operation data, which limits the speed and accuracy of data analysis and dispatching decisions. Due to the lack of effective data support and intelligent dispatching methods, traditional power grid dispatching often finds it difficult to cope with emergencies and changing supply and demand relationships, and has poor flexibility. In traditional power grids, once a fault occurs, due to the lack of real-time monitoring and rapid positioning technology, the detection and response time of the fault is long, which may lead to a wider range of power outages and affect the user's power experience. Traditional power grid dispatching may not make full use of existing power generation resources, resulting in energy waste. Especially with the increasing number of renewable energy access, how to effectively integrate various power sources and achieve the optimal energy configuration has become a challenge. Therefore, a cloud computing-based active-standby combined power grid dispatching optimization system is designed. Summary of the invention

[0003] The purpose of the present invention is to provide a cloud computing-based active-standby combined power grid dispatching optimization system to solve the problems of insufficient data processing capability, poor dispatching flexibility, slow fault response and low resource utilization proposed in the above background technology.

[0004] To achieve the above object, the present invention aims to provide a cloud computing-based active-standby combined power grid dispatch optimization system, comprising:

[0005] A data acquisition unit, which collects electrical data of the operation of the power grid from devices in the power grid and uploads the electrical data to the cloud via wireless communication;

[0006] A data storage and analysis unit, wherein the data storage and analysis unit stores the electrical data in the cloud and analyzes the electrical data using a regression model;

[0007] A scheduling decision support unit, which uses an optimal power flow algorithm to formulate a power generation plan and a load distribution strategy based on the electrical data in the data storage and analysis unit, introduces the startup time and cost of the power generation module into the objective function of the optimal power flow algorithm, and considers the influence of load demand and the availability of the power generation module over time to dynamically schedule the optimal power flow algorithm mathematical model;

[0008] The fault detection and recovery unit detects abnormal conditions in the power grid through real-time monitoring, quickly locates the fault point, and initiates an emergency plan to isolate and recover the fault.

[0009] As a further improvement of the technical solution, the electrical data of the power grid operation includes:

[0010] Voltage, current, power and frequency.

[0011] As a further improvement of the technical solution, the data storage and analysis unit uses a regression model to analyze the electrical data, including the following steps:

[0012] S1.1. Preprocess the electrical data and save them using the cloud storage service Amazon S3;

[0013] S1.2, select characteristic variables related to the operation status of the power grid, the characteristic variables include voltage, current and power factor, and use the characteristic variables as input variables;

[0014] S1.3, using historical electrical data as a training set, establish a linear regression model;

[0015] S1.4. Evaluate the performance of the linear regression model by cross-validation method and adjust the linear regression model parameters to improve the prediction accuracy;

[0016] S1.5. Apply the trained and evaluated linear regression model to real-time electrical data to predict future grid loads.

[0017] As a further improvement of the technical solution, in S1.3, the linear regression model is specifically:

[0018] ;

[0019] in, represents the target variable; represents the independent variable; represents the intercept term of the model; Indicates the independent variable The coefficient of represents the random error term.

[0020] As a further improvement of the technical solution, the optimal power flow algorithm is used to formulate a power generation plan and a load distribution strategy, including the following steps:

[0021] S2.1. Collect data related to predicted load demand, maximum and minimum output limits of power generation modules and line transmission capacity;

[0022] S2.2. According to the operation constraints and taking the ramp rate into consideration, a mathematical model of the optimal power flow is established. The objective function of the mathematical model of the optimal power flow includes minimizing the power generation cost and reducing the network loss. The startup time and cost of the power generation module are introduced into the objective function, and the influence of the load demand and the availability of the power generation module changing over time is considered to dynamically schedule the mathematical model.

[0023] S2.3. Use Newton's method to solve the established mathematical model of optimal power flow and obtain the optimal solution that meets the constraints. The optimal solution includes the optimal output value of each power generation module, the power flow in the network and the load adjustment plan;

[0024] S2.4. Based on the solution results, specific dispatching instructions are generated. The dispatching instructions include adjusting the output of the power generation module, changing the load distribution, and adjusting the state of the reactive power compensation device.

[0025] As a further improvement of the technical solution, in S2.2, the constraint conditions are specifically:

[0026] Power generation module output limit:

[0027] ;

[0028] Node voltage limit:

[0029] ;

[0030] Line flow restrictions:

[0031] ;

[0032] The climbing rate considered in the constraints is:

[0033] Upward climbing rate constraint:

[0034] ;

[0035] Downward climbing rate constraint:

[0036] ;

[0037] in, Indicates power generation module In time period Active power output; Indicates power generation module In time period -1 active power output; Indicates power generation module The upward climbing rate, that is, the maximum active power output that the power generation module can increase per unit time; Indicates a time interval; Indicates power generation module The downward ramp rate is the maximum active power output that the power generation module can reduce per unit time; Indicates power generation module In time period The minimum active power output; Indicates power generation module In time period Maximum active power output; Representation Node Minimum voltage of Representation Node Maximum voltage of Representation Node Voltage; Indicates line The minimum tidal current; Indicates line The trend of Indicates line The biggest trend.

[0038] As a further improvement of the technical solution, in S2.2, the mathematical model for establishing the optimal power flow is specifically:

[0039] ;

[0040] The startup time and cost of the power generation module are introduced into the objective function, and the mathematical model is dynamically scheduled considering the influence of load demand and power generation module changes over time. Specifically:

[0041] ;

[0042] in, Indicates active power; represents the weight coefficient for balancing the generation cost; Indicates the number of power generation modules; Represents an index; Indicates Cost function of each generator set; Indicates Active power generated by each power generation module; Indicates the weight coefficient of balancing network loss; Indicates the network loss of the entire power network; Indicates the entire power network in the time period Network loss; Indicates the total number of time periods; Indicates The power generation module in the time period The cost of electricity generation; Indicates power generation module In time period Start-up costs; Indicates the startup behavior of the power generation module.

[0043] As a further improvement of the technical solution, in S2.3, the mathematical model of the optimal power flow is solved using the Newton method, including the following steps:

[0044] S3.1. Set an initial point, i.e., the initial power generation module output value and other control variables, and set the convergence criterion as the change in the control variable is less than the threshold value p;

[0045] S3.2, calculating the first-order partial derivatives of the objective function with respect to each control variable to form a Jacobian matrix, which contains the partial derivatives of the objective function with respect to each variable;

[0046] S3.3, calculate the second-order partial derivatives of the objective function with respect to all control variables to form the Hessian matrix;

[0047] S3.4. In each iteration, the Jacobian matrix and Hessian matrix of the current point are used to estimate the quadratic model of the objective function near the point, and the direction in which the quadratic model decreases the fastest is found. Then, a step is moved along this direction to reach the new point, and the values ​​of each variable are updated.

[0048] S3.5. When the iterative process meets the pre-set convergence criteria, it means that the Newton method has converged. At this time, the output result is the optimal solution, which includes the optimal output configuration of each generator, the power flow distribution in the entire power grid, and the load adjustment strategy.

[0049] As a further improvement of the technical solution, the fault detection and recovery unit detects abnormal conditions in the power grid through real-time monitoring and quickly locates the fault point, including the following steps:

[0050] S4.1. Use the support vector machine algorithm to train historical data and establish a baseline model under normal operating conditions. When the real-time data of the power grid deviates from the normal range [ , ], trigger anomaly detection;

[0051] S4.2. Use the traveling wave method to accurately locate the fault point, and introduce the line material, atmospheric pressure and humidity into the traveling wave method to optimize the location of the fault point, and identify the fault type according to the electrical characteristics of the fault point;

[0052] S4.3. Prepare multiple emergency response plans in advance and formulate corresponding isolation and recovery measures for faults of different types and scales;

[0053] S4.4. Disconnect the faulty equipment from the grid through the remote control switch module, and adjust the grid topology to isolate the faulty area from other normal operating areas;

[0054] S4.5. While isolating the faulty area, restore power supply to areas that are less affected or not directly affected by adjusting the grid operation mode. Adjusting the grid operation mode includes reconfiguring power supply points and adjusting load distribution.

[0055] As a further improvement of the technical solution, in S4.2, the traveling wave method is specifically:

[0056] ;

[0057] ;

[0058] Introducing line material, atmospheric pressure and humidity into the traveling wave method to optimize the location of the fault point is as follows:

[0059] ;

[0060] in, It indicates the distance from the fault point to the starting point after the line material, atmospheric pressure and humidity are introduced into the traveling wave method; Indicates the distance between the fault point and the starting point; Indicates the time when the endpoint detection module records the traveling wave; Indicates the time when the starting point detection module records the traveling wave; Indicates wave speed; represents the speed of light; Indicates the line material correction factor; Indicates the atmospheric pressure correction factor; Humidity correction factor; Represents the relative dielectric constant of air.

[0061] Compared with the prior art, the present invention has the following beneficial effects:

[0062] 1. In this cloud computing-based active-standby combined power grid dispatch optimization system, by uploading real-time electrical data to the cloud and processing it using a linear regression model, the system can more accurately predict future load demand. Based on these prediction results, the system can use the optimal power flow algorithm to formulate accurate power generation plans and load distribution strategies to ensure the stability and economy of power supply. In addition, considering the startup time and cost of the power generation module, as well as the load demand and availability of the power generation module that change over time, the system can perform dynamic scheduling, further improving the overall efficiency of power grid operation.

[0063] 2. The cloud computing-based active-standby combined power grid dispatch optimization system has powerful fault detection and rapid recovery functions. With the help of real-time monitoring technology and support vector machine algorithms, the system can quickly detect abnormal conditions in the power grid and accurately locate the fault point, take isolation measures in time to prevent the fault from expanding, and restore power supply to non-faulty areas by adjusting the power grid operation mode. This rapid response mechanism not only reduces power outage time, but also reduces maintenance costs, and improves the safety and reliability of the entire power grid system. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 It is the overall flow chart of the present invention;

[0065] The meaning of each number in the figure is:

[0066] 1. Data acquisition unit; 2. Data storage and analysis unit; 3. Scheduling decision support unit; 4. Fault detection and recovery unit. DETAILED DESCRIPTION

[0067] 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. Example

[0068] See also Figure 1 As shown, a cloud computing-based active-standby combined power grid dispatch optimization system is provided, including:

[0069] The data acquisition unit 1 collects electrical data of the power grid operation from the devices in the power grid and uploads the electrical data to the cloud via wireless communication;

[0070] In this embodiment, the electrical data of the power grid operation includes:

[0071] Voltage, current, power and frequency.

[0072] The data storage and analysis unit 2 stores the electrical data in the cloud and uses a regression model to analyze the electrical data to provide a basis for power grid dispatching decisions;

[0073] In this embodiment, the regression model can be used to predict future electrical parameter values, such as load forecasting. This is crucial for the planning and operation of power systems because it can help grid operators prepare in advance to meet demand during peak hours or avoid overloads; by fitting historical data, the regression model can reveal long-term trends in electrical data, which is very important for evaluating power system performance, equipment aging, and future investment needs; based on the prediction results of the regression model, the control strategy of the power system can be optimized. Using the regression model to analyze electrical data includes the following steps:

[0074] S1.1. Preprocess the electrical data, including data cleaning and data formatting, and use the cloud storage service Amazon S3 to save the electrical data;

[0075] S1.2, select characteristic variables related to the operation status of the power grid, the characteristic variables include voltage, current and power factor, and use the characteristic variables as input variables;

[0076] S1.3, using historical electrical data as a training set, establish a linear regression model;

[0077] The linear regression model is as follows:

[0078] ;

[0079] in, represents the target variable, i.e., the grid load and power generation demand in a certain period of time in the future; Represents the independent variable, which is used to predict the dependent variable Input data, which can be any factor that affects grid load, including temperature, humidity, time (such as time of day), date (such as whether it is a weekday), and historical load data; Represents the intercept term of the model, which represents the dependent variable when all independent variables are zero. The expected value of is the basic level of grid load without the influence of other external factors; Indicates the independent variable The coefficient of the independent variable For dependent variables the extent of the impact; Represents the random error term, which represents the influencing factors that the model fails to capture.

[0080] S1.4. Evaluate the performance of the linear regression model by cross-validation method and adjust the linear regression model parameters to improve the prediction accuracy;

[0081] Among them, the cross-validation method is to divide the data into several parts during the training process, and use part of them as the validation set and the rest as the training set in turn, so that the training and validation process can be repeated many times, and finally the performance of the model can be evaluated by combining the results of multiple validations. By training and validating the model on multiple different data subsets, it can reduce the overfitting of the model to a specific training set and improve the generalization ability of the model;

[0082] S1.5. Apply the trained and evaluated linear regression model to real-time electrical data to predict future grid loads.

[0083] The dispatch decision support unit 3 formulates a power generation plan and load distribution strategy based on the electrical data in the data storage and analysis unit 2 using the optimal power flow algorithm;

[0084] In this embodiment, the optimal power flow algorithm can find the most economical power generation combination under the premise of satisfying all operating constraints, thereby minimizing the power generation cost; by optimizing the flow of electricity in the network, the loss in the power transmission process can be effectively reduced, which is crucial to improving the overall efficiency of the power system; the optimal power flow algorithm takes into account the operating constraints of the system, including the maximum and minimum output limits of the power generation module, the line transmission capacity, etc., so as to ensure that the power grid can remain stable under various operating conditions. In addition, by considering the startup time and cost of the power generation module, the power generation plan can be adjusted more flexibly to cope with emergencies and improve the reliability of the system; using the optimal power flow algorithm to formulate a power generation plan and a load distribution strategy includes the following steps:

[0085] S2.1. Collect data related to predicted load demand, maximum and minimum output limits of power generation modules and line transmission capacity;

[0086] S2.2. According to the operation constraints and taking the ramp rate into consideration in the constraints, a mathematical model of the optimal power flow is established. The objective function of the mathematical model of the optimal power flow includes minimizing the power generation cost and reducing the network loss. The startup time and cost of the power generation module are introduced into the objective function, and the influence of the load demand and the availability of the power generation module changing over time is considered to dynamically schedule the mathematical model. The power generation module represents the generator.

[0087] The specific constraints are:

[0088] Power generation module output limit:

[0089] ;

[0090] Node voltage limit:

[0091] ;

[0092] Line flow restrictions:

[0093] ;

[0094] A large change in the output of the power generation module in a short period of time may cause grid frequency fluctuations or other instability. The ramp rate limit helps to smooth the output changes of the power generation module, thereby maintaining the stable operation of the power grid; the frequent and rapid increase and decrease of the load of the power generation module will cause additional pressure on the equipment, which may cause equipment damage or shorten its service life. By limiting the ramp rate, the stress of the equipment can be reduced and its service life can be extended; the ramp rate limit provides an important constraint condition for scheduling decisions, making the scheduling plan more practical and feasible. When formulating a power generation plan, the climbing ability of the power generation module must be taken into account to ensure that the plan is feasible during actual execution; the specific climbing rate considered in the constraint condition is:

[0095] Upward climbing rate constraint:

[0096] ;

[0097] This means that the power generation module In time period The active power output cannot exceed that of the previous time period -1 active power output plus the ramp rate and time interval The product of

[0098] Downward climbing rate constraint:

[0099] ;

[0100] This means that the power generation module In time period The active power output cannot be lower than that in the previous period -1 active power output minus the ramp rate and time interval The product of

[0101] in, Indicates power generation module In time period Active power output; Indicates power generation module In time period -1 active power output; Indicates power generation module The upward climbing rate, that is, the maximum active power output that the power generation module can increase per unit time; Indicates a time interval; Indicates power generation module The downward ramp rate is the maximum active power output that the power generation module can reduce per unit time; Indicates power generation module In time period The minimum active power output; Indicates power generation module In time period Maximum active power output; Representation Node Minimum voltage of Representation Node Maximum voltage; Representation Node Voltage; Indicates line The minimum tidal current; Indicates line The trend; Indicates line The biggest trend.

[0102] Furthermore, the objective function of the mathematical model for establishing the optimal power flow algorithm is specifically:

[0103] ;

[0104] ;

[0105] The start-up time and cost of the power generation module are introduced to more accurately reflect the actual power generation cost. Since different power generation modules have different start-up costs and time requirements, these factors will affect the scheduling decision and make the scheduling plan closer to the actual situation. By introducing the start-up time and cost, the power generation cost can be optimized and unnecessary startup costs can be avoided, which helps to reduce the total power generation cost. Dynamic scheduling takes into account the influence of load demand and power generation module changes over time, making the scheduling plan more in line with the actual situation and reducing the risk of instability caused by improper scheduling. Dynamic scheduling can adjust the power generation plan according to real-time data, making the power system more flexible in responding to emergencies (such as extreme weather, equipment failure, etc.), and can quickly adjust the power generation and load distribution strategies to ensure the continuity of power supply. Considering the start-up time and cost of the power generation module, as well as the availability that changes over time, it can ensure that the scheduling plan is feasible in actual execution and improve the reliability of power grid operation. Introducing the start-up time and cost of the power generation module in the objective function and considering the influence of load demand and power generation module changes over time to dynamically schedule the mathematical model, specifically:

[0106] ;

[0107] in, Indicates active power; represents the weight coefficient for balancing the generation cost; Indicates the number of power generation modules; Represents an index; Indicates Cost function of each generator set; Indicates Active power generated by each power generation module; Indicates the weight coefficient of balancing network loss; It represents the network loss of the entire power network, that is, the loss of power during the transmission process; Indicates the entire power network in the time period Network loss; Indicates power generation module Fixed costs, which are costs that need to be paid even when no electricity is generated; is the coefficient of the first-order term, indicating the power generation model The cost of electricity generation increases with the active power The linear growth part of the increase; is the coefficient of the quadratic term, indicating the power generation module The cost of electricity generation increases with the active power The added nonlinear growth part reflects the change of efficiency with power; Indicates the total number of time periods; Indicates The power generation module in the time period The cost of electricity generation; Indicates power generation module In time period Start-up costs; Indicates the startup behavior of the power generation module.

[0108] S2.3. Use Newton's method to solve the established mathematical model of optimal power flow and obtain the optimal solution that meets the constraints. The optimal solution includes the optimal output value of each power generation module, the power flow in the network and the load adjustment plan;

[0109] Furthermore, Newton's method is a second-order optimization method that uses the first-order and second-order derivatives of the objective function (i.e., the Jacobian matrix and the Hessian matrix) to approximate the behavior of the objective function. Compared with first-order methods such as gradient descent, Newton's method can usually converge to the optimal solution faster. For nonlinear optimization problems such as the optimal power flow problem, fast convergence means that the optimal solution can be found more efficiently; Newton's method has a superlinear convergence rate when approaching the minimum point, which means that when approaching the optimal solution, the improvement of each iteration is greater, thereby speeding up the solution process. This is especially important for complex and highly nonlinear optimal power flow problems; Newton's method can better handle nonlinear constraints and can overcome the problem of local minima to a certain extent. In particular, when dealing with non-convex problems commonly found in power systems, Newton's method shows good robustness; using Newton's method to solve the established mathematical model of optimal power flow includes the following steps:

[0110] S3.1. Set an initial point, i.e., the initial power generation module output value and other control variables (including transformer tap position, reactive power compensation device, etc.), and set the convergence criterion as the change in the control variable is less than the threshold value p;

[0111] S3.2. Calculate the first-order partial derivatives of the objective function with respect to each control variable to form a Jacobian matrix. The Jacobian matrix contains the partial derivatives of the objective function with respect to each variable. The Jacobian matrix is ​​a matrix composed of the first-order partial derivatives of the objective function (including minimization of total network loss or minimization of power generation cost, etc.) with respect to each control variable (including active output and reactive output of the generator, etc.), which reflects the degree of influence of the change of the control variable on the objective function;

[0112] S3.3. Calculate the second-order partial derivatives of the objective function with respect to all control variables to form a Hessian matrix. The Hessian matrix is ​​a matrix composed of the second-order partial derivatives of the objective function with respect to all control variables. It provides information about the curvature of the objective function and helps determine the search direction.

[0113] S3.4. In each iteration, the Jacobian matrix and Hessian matrix of the current point are used to estimate the quadratic model of the objective function near the point, and the direction in which the quadratic model decreases the fastest is found. Then, a step is moved along this direction to reach the new point, and the values ​​of each variable are updated.

[0114] Among them, the construction of the secondary model is specifically as follows:

[0115] At the current point At this point, Newton's method constructs an objective function The secondary model , the model is given by the gradient and the Hessian matrix at the current point: ;in, is the Jacobian matrix (gradient), is the Hessian matrix;

[0116] Solving the minimization problem is specifically:

[0117] In order to find the quadratic model The fastest descending direction, solve the following minimization problem: ; Solve this optimization problem to get the search direction : ;

[0118] Update the control variables as follows:

[0119] Getting the search direction Then, you can move one step in this direction. Arriving at a new point : ; Step length Determined by line search method to ensure that the objective function actually decreases at the new point;

[0120] S3.5. When the iterative process meets the pre-set convergence criteria, it means that the Newton method has converged. At this time, the output result is the optimal solution, which includes the optimal output configuration of each generator, the power flow distribution in the entire power grid, and the load adjustment strategy.

[0121] S2.4. Based on the solution results, specific dispatching instructions are generated. The dispatching instructions include adjusting the output of the power generation module, changing the load distribution, and adjusting the state of the reactive power compensation device.

[0122] The fault detection and recovery unit 4 detects abnormal conditions in the power grid through real-time monitoring, quickly locates the fault point, and initiates the emergency plan to isolate and recover the fault;

[0123] In this embodiment, the fault detection and recovery unit 4 detects abnormal conditions in the power grid through real-time monitoring and quickly locates the fault point, including the following steps:

[0124] S4.1. Use the support vector machine algorithm to train historical data and establish a baseline model under normal operating conditions. When the real-time data of the power grid deviates from the normal range [ , ], trigger anomaly detection;

[0125] If there is a set of training data ,in, is the input feature vector, is the corresponding label, then the baseline model is:

[0126] ;

[0127] in, is a new input vector; represents the model's predicted value for this input; is the index of the sample in the training dataset; Represents the number of support vectors in the training dataset; represents the Lagrange multiplier, which is a non-negative real number used to weight the support vectors in the training data; Indicates the training data set The labels of the samples; represents the kernel function, which calculates the input vector and the new input vector Inner products in high-dimensional spaces; represents the bias term, which is a constant offset of the decision boundary and determines the position of the hyperplane;

[0128] Further , ; It represents the mean of the features closely related to the real-time data of the power grid, including voltage, current, power factor, frequency, represents the standard deviation, Indicates experience value;

[0129] If the voltage amplitude of the grid node is 100V and the standard deviation is 5V under normal operation, set =2 (i.e. twice the standard deviation), then the normal range [ , ] is defined as:

[0130] ;

[0131] ;

[0132] This means that if the voltage amplitude monitored in real time exceeds the range of 90V to 110V, it is considered that the node may be abnormal, triggering anomaly detection;

[0133] S4.2. Use the traveling wave method to accurately locate the fault point, and introduce the line material, atmospheric pressure and humidity into the traveling wave method to optimize the location of the fault point. According to the electrical characteristics of the fault point (including current waveform, voltage drop, etc.), the fault type is identified, including single-phase grounding fault, two-phase short circuit fault and three-phase short circuit fault;

[0134] Among them, the traveling wave method can quickly detect the moment when the fault occurs, and determine the location of the fault point by analyzing the time difference of the traveling wave propagation. This process is completed almost in real time, which greatly shortens the time of fault location and helps to take measures quickly for repair; by measuring the time difference of the traveling wave propagating from the fault point to the detection point, and combining parameters such as wave speed, the specific location of the fault point can be accurately calculated. This method can achieve high accuracy in theory, especially when locating faults within a short distance; the traveling wave method is applicable to various power grid structures, whether it is a simple radial network or a complex ring network, it can be effectively applied, because the propagation path of the traveling wave signal in the power grid is clear and is not affected by the topological structure of the power grid; the traveling wave method is specifically:

[0135] ;

[0136] ;

[0137] The purpose of introducing line material, atmospheric pressure and humidity parameters is to achieve more refined power grid management. By considering more factors that affect the propagation of traveling waves, the intelligence level of the fault detection and positioning system can be improved. Different line materials have different resistivities. Materials with high resistivity will increase the energy loss of traveling waves during transmission, thereby affecting the speed of traveling waves. The dielectric constant of line materials will also affect the propagation speed of traveling waves. Materials with high dielectric constants will slow down the propagation speed of traveling waves. Different materials have different dielectric constants, so they will affect the propagation speed of traveling waves. Changes in atmospheric pressure directly affect the density of air. An increase in air density will increase the probability of collision between air molecules, thereby increasing the propagation resistance of traveling waves in the air, resulting in a higher propagation speed of traveling waves. Slow down; the increase in humidity will lead to an increase in the water vapor content in the air, and the dielectric constant of water is much higher than the dielectric constant of dry air. Therefore, the increase in humidity will increase the dielectric constant of the air, thereby slowing down the propagation speed of the traveling wave; since environmental factors will cause changes in the speed of the traveling wave, if these factors are not taken into account, it may cause positioning deviations. After introducing these parameters, it can reduce misjudgments caused by environmental changes and improve the reliability of positioning; in different geographical environments, line materials, atmospheric pressure, humidity and other conditions may vary significantly. After introducing these parameters, the traveling wave method can better adapt to various complex environments and ensure that the fault point can be accurately located under different conditions; introducing line material, atmospheric pressure and humidity into the traveling wave method to optimize the location of the fault point is as follows:

[0138] ;

[0139] in, It indicates the distance from the fault point to the starting point after the line material, atmospheric pressure and humidity are introduced into the traveling wave method; Indicates the distance between the fault point and the starting point; Indicates the time when the endpoint detection module records the traveling wave; Indicates the time when the starting point detection module records the traveling wave; Indicates wave speed; represents the speed of light; Indicates the line material correction factor; Indicates the atmospheric pressure correction factor; Humidity correction factor; represents the relative dielectric constant of air;

[0140] S4.3. Prepare multiple emergency plans in advance, formulate corresponding isolation and recovery measures for faults of different types and scales, and set up a main plan and multiple backup plans in the system. When the main plan is unavailable, it can quickly switch to the backup plan;

[0141] Among them, a variety of emergency plans are formulated to formulate corresponding isolation and recovery measures for faults of different types and scales, including the following steps:

[0142] If it is a small-scale local fault, the fault type is single-phase ground short circuit, local line break, etc., the isolation measures are: use the remote control switch module to automatically disconnect the faulty equipment from the grid or adjust the grid topology to isolate the faulty area from other normal operating areas; the recovery measures are: reallocate the load in the affected area to other available power supply points, enable backup power sources (such as backup generators) to supply power to critical loads, or gradually restore power supply, giving priority to important users and critical infrastructure;

[0143] If it is a large-scale system failure, the fault type is multi-phase short circuit, large-scale line break, etc., then the isolation measures are: the system quickly disconnects multiple fault points to prevent the fault from spreading or adjusts the grid structure to minimize the scope of the fault; the recovery measures are: give priority to restoring the power supply to important users, such as hospitals, data centers, etc., and gradually restore the power supply in stages, first restore the areas with less impact on the stability of the grid, adjust the load distribution, and ensure the stability of the grid during the recovery process; S4.4, disconnect the faulty equipment from the grid through the remote control switch module to prevent the fault from spreading, and adjust the grid topology to isolate the faulty area from other normal operating areas. This is equivalent to quickly switching to the backup path or backup equipment when a problem occurs in the main system to ensure the normal operation of other parts;

[0144] S4.5. While isolating the faulty area, the power supply to areas that are less affected or not directly affected is restored by adjusting the grid operation mode. Adjusting the grid operation mode includes reconfiguring power supply points and adjusting load distribution. This means that the system can not only quickly isolate the fault when it occurs, but also restore power supply by adjusting backup resources, thereby ensuring the overall stability and reliability of the system.

[0145] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments, and the above embodiments and descriptions are only preferred examples of the present invention, and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, and these changes and improvements all fall within the scope of the present invention to be protected.

Claims

1. A cloud computing-based active-standby combined power grid dispatching optimization system, characterized in that: include: A data collection unit (1), wherein the data collection unit (1) collects electrical data of power grid operation from devices in the power grid, and uploads the electrical data to the cloud via wireless communication; A data storage and analysis unit (2), wherein the data storage and analysis unit (2) stores the electrical data in the cloud and uses a regression model to analyze the electrical data; A scheduling decision support unit (3), wherein the scheduling decision support unit (3) uses an optimal power flow algorithm based on the electrical data in the data storage and analysis unit (2) to formulate a power generation plan and a load distribution strategy, introduces the startup time and cost of the power generation module into the objective function of the optimal power flow algorithm, and considers the influence of the load demand and the availability of the power generation module changing over time to dynamically schedule the optimal power flow algorithm mathematical model; A fault detection and recovery unit (4), wherein the fault detection and recovery unit (4) detects abnormal conditions in the power grid through real-time monitoring, quickly locates the fault point, and initiates an emergency plan to isolate and recover the fault; The method of adopting the optimal power flow algorithm to formulate a power generation plan and a load distribution strategy includes the following steps: S2.

1. Collect data related to predicted load demand, maximum and minimum output limits of power generation modules and line transmission capacity; S2.

2. According to the operation constraints and taking the ramp rate into consideration, a mathematical model of the optimal power flow is established. The objective function of the mathematical model of the optimal power flow includes minimizing the power generation cost and reducing the network loss. The startup time and cost of the power generation module are introduced into the objective function, and the influence of the load demand and the availability of the power generation module changing over time is considered to dynamically schedule the mathematical model. S2.

3. Use Newton's method to solve the established mathematical model of optimal power flow and obtain the optimal solution that meets the constraints. The optimal solution includes the optimal output value of each power generation module, the power flow in the network and the load adjustment plan; S2.

4. Based on the solution results, generate specific dispatching instructions, which include adjusting the output of the power generation module, changing the load distribution and adjusting the state of the reactive power compensation device; In S2.2, the constraints are specifically: Power generation module output limit: ； Node voltage limit: ； Line flow restrictions: ； The climbing rate considered in the constraints is: Upward climbing rate constraint: ； Downward climbing rate constraint: ； in, Indicates power generation module In time period Active power output; Indicates power generation module In time period Active power output; Indicates power generation module The upward climbing rate, that is, the maximum active power output that the power generation module can increase per unit time; Indicates a time interval; Indicates power generation module The downward ramp rate is the maximum active power output that the power generation module can reduce per unit time; Indicates power generation module In time period The minimum active power output; Indicates power generation module In time period Maximum active power output; Representation Node Minimum voltage of Representation Node Maximum voltage; Representation Node Voltage; Indicates line The minimum tidal current; Indicates the power flow of the line; Indicates line The biggest trend; In S2.2, the mathematical model for establishing the optimal power flow is specifically: ； The startup time and cost of the power generation module are introduced into the objective function, and the mathematical model is dynamically scheduled considering the influence of load demand and power generation module changes over time. Specifically: ； in, Indicates active power; The weight coefficient representing the balanced generation cost; Indicates the number of power generation modules; Represents an index; Indicates Cost function of each generator set; Indicates Active power generated by each power generation module; Indicates the weight coefficient of balancing network loss; Indicates the network loss of the entire power network; Represents the entire power network in the time period Network loss; Indicates the total number of time periods; Indicates The power generation module in the time period The cost of electricity generation; Indicates power generation module In time period Start-up costs; Indicates the startup behavior of the power generation module; In S2.3, the Newton method is used to solve the established mathematical model of the optimal power flow, including the following steps: S3.

1. Set an initial point, i.e., the initial power generation module output value and other control variables, and set the convergence criterion as the change in the control variable is less than the threshold value p; S3.2, calculating the first-order partial derivatives of the objective function with respect to each control variable to form a Jacobian matrix, which contains the partial derivatives of the objective function with respect to each variable; S3.3, calculate the second-order partial derivatives of the objective function with respect to all control variables to form the Hessian matrix; S3.

4. In each iteration, the Jacobian matrix and Hessian matrix of the current point are used to estimate the quadratic model of the objective function near the point, and the direction in which the quadratic model decreases the fastest is found. Then, a step is moved along this direction to reach the new point, and the values ​​of each variable are updated. S3.

5. When the iterative process meets the pre-set convergence criteria, it means that the Newton method has converged. At this time, the output result is the optimal solution.

2. The cloud computing-based active-standby combined power grid dispatch optimization system according to claim 1 is characterized in that: The electrical data of the power grid operation includes voltage, current, power and frequency.

3. The cloud computing-based active-standby combined power grid dispatch optimization system according to claim 1 is characterized in that: The data storage and analysis unit (2) uses a regression model to analyze the electrical data, including the following steps: S1.

1. Preprocess the electrical data and save them using the cloud storage service Amazon S3; S1.2, select characteristic variables related to the operation status of the power grid, the characteristic variables include voltage, current and power factor, and use the characteristic variables as input variables; S1.3, using historical electrical data as a training set, establish a linear regression model; S1.

4. Evaluate the performance of the linear regression model by cross-validation method and adjust the linear regression model parameters to improve the prediction accuracy; S1.

5. Apply the trained and evaluated linear regression model to real-time electrical data to predict future grid loads.

4. The cloud computing-based active-standby combined power grid dispatch optimization system according to claim 3 is characterized by: In S1.3, the linear regression model is specifically: ； in, represents the target variable; represents the independent variable; represents the intercept term of the model; Indicates the independent variable The coefficient of represents the random error term.

5. The cloud computing-based active-standby combined power grid dispatch optimization system according to claim 1, characterized in that: The fault detection and recovery unit (4) detects abnormal conditions in the power grid through real-time monitoring and quickly locates the fault point, including the following steps: S4.

1. Use the support vector machine algorithm to train historical data and establish a baseline model under normal operating conditions. When the real-time data of the power grid deviates from the normal range When , anomaly detection is triggered; S4.

2. Use the traveling wave method to accurately locate the fault point, and introduce the line material, atmospheric pressure and humidity into the traveling wave method to optimize the location of the fault point, and identify the fault type according to the electrical characteristics of the fault point; S4.

3. Prepare multiple emergency response plans in advance and formulate corresponding isolation and recovery measures for faults of different types and scales; S4.

4. Disconnect the faulty equipment from the grid through the remote control switch module, and adjust the grid topology to isolate the faulty area from other normal operating areas; S4.

5. While isolating the faulty area, restore power supply to areas that are less affected or not directly affected by adjusting the grid operation mode. Adjusting the grid operation mode includes reconfiguring power supply points and adjusting load distribution.

6. The cloud computing-based active-standby combined power grid dispatch optimization system according to claim 5 is characterized by: In S4.2, the traveling wave method is specifically: ； ； Introducing line material, atmospheric pressure and humidity into the traveling wave method to optimize the location of the fault point is as follows: ； in, It indicates the distance from the fault point to the starting point after the line material, atmospheric pressure and humidity are introduced into the traveling wave method; Indicates the distance between the fault point and the starting point; Indicates the time when the endpoint detection module records the traveling wave; Indicates the time when the starting point detection module records the traveling wave; Indicates wave speed; represents the speed of light; Indicates the line material correction factor; Indicates the atmospheric pressure correction factor; Humidity correction factor; Represents the relative dielectric constant of air.

Images