Bus energy balance adjusting system

Abstract

The invention discloses a bus energy balance adjustment system, and relates to the technical field of bus balance, the bus energy balance adjustment system comprises a system clock module, a data acquisition module, a three-phase balance module, a threshold adjustment module, a balance identification module, a maintenance correction module and a cloud sharing module, the system clock module sets nodes and periods, and the data acquisition module acquires data required by the system; the three-phase balance module adopts an intelligent reactive compensation method to realize dynamic three-phase balance; the threshold adjustment module considers the three-phase voltage unbalance degree and the temperature, and determines a dynamic bus abnormal threshold based on a kernel density estimation method in combination with a bus line loss rate; the balance identification module adopts a balance analysis model and a progressive traceability strategy to decide whether to send a branching number and / or a line loss error sequence; the maintenance correction module generates a branching ammeter replacement prompt based on the branching number, and generates a bus adjustment coefficient and a branching adjustment coefficient vector of a next period based on a load regulation strategy; and the cloud sharing module realizes data sharing of each module in the system.

Description

Technical Field

[0001] The present invention relates to the technical field of busbar balancing, and in particular to a busbar energy balancing regulation system. Background Art

[0002] The bus energy imbalance rate is an important technical indicator that reflects the bus line loss of each voltage level of the substation and measures the power supply and output of the substation. It is also an intuitive reflection of the power loss in the station. The electric energy collection system uses the measurement and conversion of the voltage and current transformers of the primary equipment in the station, connects to the electric energy meter with the secondary line, and then transmits the measured data from the plant to the main station through the communication equipment, so as to realize the measurement of the power in and out of the bus at all levels in the station and the statistics of the bus imbalance rate. By monitoring the bus imbalance rate, it is helpful to timely discover the bus balance fault, eliminate the defect as soon as possible, and realize the scientific and economic management of the power grid.

[0003] The existing Chinese patent with authorization announcement number CN114792975B proposes a data repair method for bus balancing in a power grid, including: obtaining the inflow electric energy and outflow electric energy of the meter on the bus in any selected power line group; establishing a bus balance relationship; establishing a branch line balance relationship; establishing a meter restriction relationship; using the bus balance relationship, the branch line balance relationship and the meter restriction relationship to solve the first adjustment coefficient and the second adjustment coefficient of the meter on the bus; and using the solved first adjustment coefficient and the second adjustment coefficient to adjust the inflow electric energy and / or outflow electric energy of the meter on the bus to balance the bus in the selected power line group; so as to realize automatic bus balance repair and improve repair efficiency and accuracy.

[0004] However, the existing technology only performs real-time data repair for bus energy imbalance. As the bus ages during use, the existing technology needs to be frequently repaired or adjusted. Therefore, the existing technology lacks effective automated analysis means and prevention means. Automated analysis means help to analyze and trace the causes of bus energy imbalance and deal with it in a timely manner, and prevention means can effectively reduce the probability of bus energy imbalance. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention proposes a busbar energy balance regulation system to realize a multifunctional and efficient regulation system integrating real-time regulation, source tracing and pre-emptive prevention.

[0006] The technical solution to achieve the purpose of the present invention is:

[0007] A busbar energy balance regulation system includes a system clock module, a data acquisition module, a three-phase balance module, a threshold adjustment module, a balance identification module, a maintenance correction module and a cloud sharing module;

[0008] The system clock module sets the total number of nodes and single-cycle nodes N, divides the consecutive N nodes into the same cycle, and announces the node settings and cycle division;

[0009] The data collection module collects the data required by the system at each node and uploads it to the cloud sharing module;

[0010] The three-phase balance module uses intelligent reactive power compensation method to achieve dynamic three-phase balance of power supply at a single node, avoiding voltage loss in the electric energy meter on the main transformer side;

[0011] The threshold adjustment module calls the three-phase voltage sequence and temperature sequence of a single cycle and transforms them into a type index sequence. The kernel density estimation method is combined with the bus line loss rate of a single cycle to obtain the predicted bus line loss rate distribution of the next cycle and determine the dynamic bus abnormality threshold of the next cycle.

[0012] The balance identification module calls the power sequence of the main transformer side and the branch line side in a single cycle, calculates and obtains the line loss error sequence of a single cycle in combination with the balance analysis model, and decides whether to send the branch line number and / or line loss error sequence to the maintenance correction module based on the progressive tracing strategy;

[0013] The maintenance correction module obtains the branch line number and integrates it to generate a branch line meter replacement prompt, obtains the line loss error sequence of a single cycle, and generates the busbar regulation coefficient and branch line regulation coefficient vector of the next cycle based on the load control strategy;

[0014] The cloud sharing module stores fixed thresholds and uploaded data from each module, and responds to data calls from each module in real time.

[0015] Furthermore, the three-phase balancing module uses an intelligent reactive power compensation method to achieve dynamic three-phase balancing of power at a single node, including the following specific steps:

[0016] Get the nth node in the ith cycle The three-phase voltage vector and three-phase current vector Three-phase voltage vector Including phase a voltage b-phase voltage and c phase voltage Three-phase current vector Including a phase current b-phase current and c-phase current N is the total number of nodes in a single cycle, and the voltage of phase a b-phase voltage and c phase voltage The mutual representation can be performed based on the phase offset α;

[0017] Obtaining the load admittance of phase ab based on the triangle load balancing formula bc phase load admittance and ca phase load admittance And further calculate the positive sequence current and negative sequence current To obtain the current imbalance

[0018] Call current imbalance threshold ζ I , determine the current imbalance Is it greater than the current imbalance threshold ζ I , when the current unbalance Greater than the current imbalance threshold ζ I When the positive sequence current is compensated, the symmetrical component principle is used to confirm the positive sequence current compensation. and compensate negative sequence current

[0019] Based on the triangle compensation balance formula combined with compensation positive sequence current and compensate negative sequence current Derived ab phase compensation admittance vector bc phase compensation admittance vector and ca phase compensation admittance vector

[0020] Further, the threshold adjustment module includes an index prediction unit, a fluctuation fitting unit, a distribution approximation unit and an analysis and decision unit;

[0021] The index prediction unit generates a single-cycle type index sequence, obtains the predicted type index sequence of the next cycle through LSTM, and uploads it to the cloud sharing module;

[0022] The fluctuation fitting unit uses the kernel density estimation method to fit the forecast type index sequence of the next period into the forecast type index distribution and uploads it to the cloud sharing module;

[0023] The distribution approximation unit uses the forecast type exponential distribution of the next cycle to adjust the bus line loss rate of a single cycle, obtains the forecast bus line loss rate distribution of the next cycle and uploads it to the cloud sharing module;

[0024] The analysis and decision-making unit determines the dynamic bus abnormality threshold for the next cycle with reference to the predicted bus line loss rate distribution for the next cycle and uploads it to the cloud sharing module.

[0025] Furthermore, the index prediction unit generates a type index sequence of a single cycle and obtains a predicted type index sequence of the next cycle, including the following specific steps:

[0026] Call the three-phase voltage sequence U(i) of the i-th cycle and confirm the n-th node The three-phase voltage vector The maximum phase voltage in and minimum phase voltage Calculating voltage unbalance

[0027] Sort N nodes The voltage unbalance degree constructs the voltage unbalance degree sequence d u (i), call the temperature sequence And predict bus line loss rate distribution

[0028] On N nodes The above prediction of bus line loss rate distribution Resample to obtain the predicted bus line loss rate sequence of the i-th cycle

[0029] Calculate the voltage line loss correlation coefficient ω of the i-th cycle respectively dl And the temperature line loss correlation coefficient ω hl , and are used as voltage unbalance degree sequence d u (i) and the weight of the temperature sequence h(i), and generate the type index sequence λ through weighted sum dh (i);

[0030] The type index sequence λ dh (i) Input LSTM to capture trend, periodicity and volatility, and generate the forecast type index sequence for the i+1th period

[0031] Furthermore, the fluctuation fitting unit adopts the kernel density estimation method to fit the prediction type index sequence of the next period into the prediction type index distribution, including the following specific steps:

[0032] Call the prediction type index sequence for the i+1th period Select Gaussian function as the kernel function K(x) for kernel density estimation;

[0033] Determine the prediction bandwidth of the kernel function K(x) in the i+1th cycle based on the Silverman rule of thumb

[0034] Based on kernel function K(x) and predicted bandwidth Get the forecast type exponential distribution for the i+1th period

[0035] Furthermore, the distribution approximation unit obtains the predicted bus line loss rate distribution of the next cycle, including the following specific steps:

[0036] Call the bus line loss rate l of the i-th cycle i As the predicted bus line loss rate distribution of the i+1th cycle The predicted distribution line loss mean

[0037] Call the forecast type exponential distribution for the i+1th period Calculate the standard deviation of the forecast distribution index And as the predicted bus line loss rate distribution of the i+1th cycle The standard deviation of the predicted distribution line loss

[0038] Set up a standard normal distribution to predict the mean of the distribution line loss and the standard deviation of the predicted distribution line loss Make corrections to obtain the predicted bus line loss rate distribution of the i+1th cycle

[0039] Furthermore, the analysis decision unit adaptively determines the dynamic bus abnormality threshold of the next cycle, including the following specific steps:

[0040] Call the predicted bus line loss rate distribution of the i+1th cycle Calculate the cumulative distribution of the predicted bus line loss rate in the i+1th cycle by integration

[0041] Get the predicted bus line loss rate cumulative distribution The predicted bus line loss rate corresponding to a value less than or equal to 0.8 is used as the dynamic bus abnormality threshold of the i+1th cycle.

[0042] Further, the balance identification module includes an equation measurement unit and an identification decision unit;

[0043] The equation measurement unit calls the power sequence of the main transformer side and the power sequence of the branch line side in a single cycle, calculates the line loss error sequence of a single cycle in combination with the balance analysis model and uploads it to the cloud sharing module;

[0044] The identification decision unit independently decides whether to send the branch line number and / or line loss error sequence to the maintenance correction module based on the progressive tracing strategy.

[0045] Furthermore, the equation measurement unit combines the balance analysis model to calculate the line loss error sequence of a single cycle, including the following specific steps:

[0046] Call the main transformer side power sequence of the i-th cycle And the power sequence of the branch side

[0047] Construct a balance analysis equation to describe the balance relationship between the total power on the main transformer side and the total power on the branch line side of the bus at a single node;

[0048] Combine the equilibrium analysis equations of N nodes in the ith period and construct the equilibrium analysis model, and further introduce the simplified matrix A of the ith period i Solution vector x i Make a simplified representation;

[0049] Solve the equilibrium analysis model based on linear algebra knowledge to obtain the solution vector x i , and transpose the solution vector (x i ) T The first M solutions are subtracted by 1 and then the inverse is taken to convert them into the line loss error sequence of the i-th cycle.

[0050] Furthermore, the identification decision unit autonomously decides whether to send the branch line number and / or line loss error sequence to the maintenance correction module based on the progressive tracing strategy, including the following specific steps:

[0051] Call measurement error rate threshold ξ β , the line loss error sequence of the i-th cycle and dynamic bus abnormality threshold

[0052] Determine line loss error sequence The bus line loss rate l i Is it greater than the dynamic bus abnormality threshold?

[0053] If it is greater than the dynamic bus abnormal threshold Further determine whether the measurement error rate of each branch line in the i-th period is greater than the measurement error rate threshold ξ β ;

[0054] If there is a measurement error rate of the mth branch line in the i-th cycle Greater than the measurement error rate threshold ξ β , send the branch line number m to the maintenance and correction module, m = 1, ..., M, M is the total number of branch lines;

[0055] If the measurement error rate of each branch line in the i-th cycle is less than or equal to the measurement error rate threshold ξ β , sending line loss error sequence To the maintenance and correction module;

[0056] If it is less than the dynamic bus abnormal threshold Further determine whether the measurement error rate of each branch line in the i-th period is greater than the measurement error rate threshold ξβ ;

[0057] If there is a measurement error rate of the mth branch line in the i-th cycle Greater than the measurement error rate threshold ξ β , send the branch line number m to the maintenance and correction module.

[0058] Further, the maintenance correction module includes an electric meter maintenance unit, a coefficient calculation unit and an adjustment execution unit;

[0059] The maintenance correction module integrates the branch line numbers received in a single cycle to generate branch line meter replacement prompts;

[0060] The coefficient calculation unit obtains the line loss error sequence of a single cycle, generates the busbar regulation coefficient and branch line regulation coefficient vector of the next cycle based on the load regulation strategy, and sends them to the regulation execution unit;

[0061] The regulation execution unit performs regulation at each node in the next cycle based on the bus regulation coefficient and the branch regulation coefficient vector.

[0062] Furthermore, the coefficient calculation unit generates the bus regulation coefficient and branch line regulation coefficient vector of the next cycle based on the load regulation strategy, including the following specific steps:

[0063] Get the line loss error sequence of the i-th cycle And call the main transformer side power sequence Q of the i-th cycle in (i) and the power sequence Q on the branch side out (i);

[0064] Establish the nth node of the i-th cycle The bus regulation relational expression of N nodes is combined to generate a bus regulation model;

[0065] Based on the maximum total power Q of the main transformer side called in,max , the minimum total power on the main transformer side Q in,min , Maximum power of branch line q out,max And the minimum power of the branch line q out,min Set up the power regulation constraint group;

[0066] The right side of the bus regulation model equation is defined as the bus regulation coefficient γ for the i+1th cycle i+1 and M branch line adjustment coefficients The busbar regulation function The optimization objective is defined as minimizing the bus regulation function, and the regulation power constraint group is used as the constraint condition of the optimization problem;

[0067] The Lagrange multiplier method is used to introduce the constraints into the optimization objective, and the busbar adjustment coefficient γ of the i+1th cycle is obtained by solving the KKT condition.i+1 and M branch line adjustment coefficients And combine M branch line adjustment coefficients Construct the line adjustment coefficient vector η i+1 .

[0068] Compared with the prior art, the present invention has the following significant advantages:

[0069] 1. Design a three-phase balance module and use intelligent reactive power compensation method to achieve dynamic three-phase balance of power supply at a single node, avoid voltage loss of the electric energy meter on the main transformer side, and reduce the probability of bus energy imbalance from the power supply side;

[0070] 2. The threshold adjustment module is designed to comprehensively consider the impact of three-phase voltage imbalance and temperature on bus line loss rate, uniformly generate type index sequence, and use kernel density estimation method to generate type index distribution. Combined with the bus line loss rate of a single cycle, the predicted bus line loss rate distribution of the next cycle is obtained to adaptively determine the dynamic bus abnormality threshold of the next cycle, avoiding the missed detection problem caused by the fixed threshold in the existing technology;

[0071] 3. Design the linkage between the balance identification module and the maintenance correction module. The balance identification module combines the balance analysis model and the progressive traceability strategy to independently judge the bus energy balance and the branch meter measurement error. When the bus energy is unbalanced, it will find out the cause of the imbalance in time. The maintenance correction module will maintain and replace the branch meter to prevent and adjust the bus energy imbalance caused by the measurement error. Based on the load control strategy, the power on the main transformer side and the branch side of the bus will be adjusted to meet the bus energy balance. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] Figure 1 It is a schematic diagram of a busbar energy balance regulation system;

[0073] Figure 2 A flow chart of obtaining a prediction type index sequence for an index prediction unit in the present invention;

[0074] Figure 3 This is a flow chart of the progressive tracing strategy in the present invention;

[0075] Figure 4 This is a flow chart of the load control strategy in the present invention. DETAILED DESCRIPTION

[0076] The present invention is further described in detail below in conjunction with the accompanying drawings and embodiments.

[0077] Example 1

[0078] like Figure 1As shown, a specific embodiment of the present invention discloses a bus energy balance regulation system, including a system clock module, a data acquisition module, a three-phase balance module, a threshold adjustment module, a balance identification module, a maintenance correction module and a cloud sharing module;

[0079] The system clock module evenly sets the total number of nodes and single-cycle nodes N, divides the consecutive N nodes into the same cycle, and announces the node settings and cycle divisions to each module in the system to achieve system time synchronization;

[0080] The data acquisition module collects the data required by the system at each node and uploads it to the cloud sharing module. The data required by the system include three-phase voltage vector, three-phase current vector, temperature, total power on the main transformer side and power vector on the branch line side. The three-phase voltage vector and three-phase current vector are obtained through the voltage meter and current meter installed at the power supply end, the temperature is obtained through the temperature sensor, the total power on the main transformer side is obtained through the busbar gateway meter installed at the busbar gateway, and the power vector on the branch line is obtained through the branch line meter installed on each branch line;

[0081] The three-phase balance module uses intelligent reactive power compensation method to achieve dynamic three-phase balance of power supply at a single node, avoiding voltage loss in the electric energy meter on the main transformer side;

[0082] The threshold adjustment module calls the three-phase voltage sequence and temperature sequence at the end node of a single cycle and transforms them into a type index sequence. The kernel density estimation method is used to generate the type index distribution. The predicted bus line loss rate distribution of the next cycle is obtained by combining the bus line loss rate of a single cycle and the dynamic bus abnormality threshold of the next cycle is adaptively determined.

[0083] The balance identification module calls the power sequence on the main transformer side and the power sequence on the branch line side at the end node of a single cycle, calculates the line loss error sequence of a single cycle in combination with the balance analysis model, and independently decides whether to send the branch line number and / or line loss error sequence to the maintenance correction module based on the progressive tracing strategy;

[0084] The maintenance correction module obtains the branch line number and integrates it to generate a branch line meter replacement prompt, obtains the line loss error sequence of a single cycle, and generates the busbar regulation coefficient and branch line regulation coefficient vector of the next cycle based on the load control strategy;

[0085] The cloud sharing module stores fixed thresholds in advance, stores them according to the nodes and cycles of the uploaded data, and classifies the data of each node in the same cycle into corresponding data sequences, responding to the data calls of other modules in real time. The fixed thresholds include the current imbalance threshold ξ I , measurement error rate threshold ξ β , Maximum total power on the main transformer side Q in,max , the minimum total power on the main transformer side Q in,min , Maximum power of branch line qout,max And the minimum power of the branch line q out,min .

[0086] Furthermore, the three-phase balancing module uses an intelligent reactive power compensation method to achieve dynamic three-phase balancing of power at a single node, including the following specific steps:

[0087] Call the nth node in the ith cycle in the cloud sharing module The three-phase voltage vector and three-phase current vector in, and The nth node in the ith cycle is The a-phase voltage, b-phase voltage and c-phase voltage, and The nth node in the ith cycle is a-phase current, b-phase current and c-phase current, n=1,…,N, N is the total number of nodes in a single cycle, due to the a-phase voltage b-phase voltage and c phase voltage The phase difference between any two phases is 120 degrees, and the voltage of phase A is b-phase voltage and c phase voltage They can be expressed mutually, and the specific formula is as follows:

[0088]

[0089] Where α represents the phase shift, j represents an imaginary number, j 2 = -1;

[0090] Get the nth node in the i-th cycle based on the triangle load balancing formula The ab phase load admittance bc phase load admittance and ca phase load admittance The triangle load balancing formula is as follows

[0091]

[0092] Based on the nth node in the ith cycle Positive sequence current and negative sequence current Calculate the nth node in the i-th cycle Current imbalance Among them, the positive sequence current and negative sequence current Through the ab phase load admittance bc phase load admittance and ca phase load admittance It is deduced that the derivation formula is as follows

[0093]

[0094] Call the current imbalance threshold ζ from the cloud sharing module I , determine the nth node in the i-th cycle Current imbalance Is it greater than the current imbalance threshold ζ I If the current imbalance Less than or equal to the current imbalance threshold ζ I , the three-phase power supply is in a three-phase load balance state. If the current is unbalanced Greater than the current imbalance threshold ζ I , the three-phase power supply is in a three-phase load unbalanced state;

[0095] Confirm the nth node in the i-th cycle based on the principle of symmetrical components Compensation positive sequence current and compensate negative sequence current Im(·) represents the operation of taking the imaginary part. The principle of symmetrical components requires that the imaginary part of the positive sequence current and negative sequence current are all 0, so the compensation positive sequence current and compensate negative sequence current Need to take the opposite number;

[0096] Based on the triangle compensation equalization formula, the nth node in the ith cycle is derived The ab phase compensation admittance vector bc phase compensation admittance vector and ca phase compensation admittance vector The triangle compensation balance formula is as follows

[0097]

[0098] Further, the threshold adjustment module includes an index prediction unit, a fluctuation fitting unit, a distribution approximation unit and an analysis and decision unit;

[0099] The index prediction unit comprehensively considers the correlation between voltage imbalance and temperature and bus line loss rate to generate a single-cycle type index sequence, and obtains the predicted type index sequence of the next cycle through LSTM and uploads it to the cloud sharing module;

[0100] The fluctuation fitting unit uses the kernel density estimation method to fit the forecast type index sequence of the next period into the forecast type index distribution and uploads it to the cloud sharing module;

[0101] The distribution approximation unit uses the forecast type exponential distribution of the next cycle to adjust the bus line loss rate of a single cycle, obtains the forecast bus line loss rate distribution of the next cycle and uploads it to the cloud sharing module;

[0102] The analysis and decision-making unit refers to the predicted bus line loss rate distribution of the next cycle to adaptively determine the dynamic bus abnormality threshold of the next cycle and upload it to the cloud sharing module.

[0103] like Figure 2 As shown, further, the index prediction unit generates a type index sequence of a single cycle and obtains a predicted type index sequence of the next cycle, including the following specific steps:

[0104] Call the three-phase voltage sequence of the i-th cycle in the cloud sharing module in, is the nth node in the ith cycle The three-phase voltage vector, n = 1, ..., N, N is the total number of nodes in a single cycle;

[0105] Confirm the nth node in the i-th cycle The three-phase voltage vector The maximum phase voltage in and minimum phase voltage Calculate the nth node in the i-th cycle Voltage unbalance The specific calculation formula is as follows:

[0106]

[0107] in, is the nth node in the ith cycle The average phase voltage of the sequence;

[0108] Sort the N nodes in the i-th cycle The voltage unbalance degree of the i-th cycle is constructed by And further call the temperature sequence of the i-th cycle in the cloud sharing module And predict bus line loss rate distribution Among them, the bus line loss rate distribution of the i-th cycle is It is generated by the distribution approximation unit at the end node of the i-1th cycle;

[0109] In the i-th cycle, N nodes The distribution of bus line loss rate for the i-th cycle is Resample to obtain the predicted bus line loss rate sequence of the i-th cycle in, is the nth node in the ith cycle The predicted bus line loss rate;

[0110] Calculate the voltage line loss correlation coefficient ω of the i-th cycle respectively dl And the temperature line loss correlation coefficient ω hl , the specific calculation formula is as follows:

[0111]

[0112] in, and are the mean value of the sequence voltage imbalance, the mean value of the sequence temperature and the mean value of the predicted sequence line loss in the i-th cycle respectively;

[0113] The voltage line loss correlation coefficient ω of the i-th cycle dl And the temperature line loss correlation coefficient ω hl As the voltage unbalance degree sequence d u (i) and the weight of the temperature sequence h(i), and generate the type index sequence through weighted sum in, is the nth node in the ith cycle The type index, type index sequence λ dh The specific calculation formula of (i) is dh (i) = ω dl ·d u (i)+ω hl h(i);

[0114] The type index sequence λ of the i-th period dh (i) Input LSTM, LSTM captures the type index sequence λ dh (i) The trend, periodicity and volatility of the forecast type index sequence of the i+1th period are generated The specific form is in, The nth node in the i+1th cycle The prediction type index.

[0115] Furthermore, the fluctuation fitting unit adopts the kernel density estimation method to fit the prediction type index sequence of the next period into the prediction type index distribution, including the following specific steps:

[0116] Call the prediction type index sequence of the i+1th period from the cloud sharing module The Gaussian function is selected as the kernel function K(x) of kernel density estimation. The specific formula of the kernel function is as follows:

[0117]

[0118] Determine the prediction bandwidth of the kernel function K(x) in the i+1th cycle based on the Silverman rule of thumb in, is the prediction type index sequence of the i+1th period The standard deviation of the predicted series index;

[0119] Prediction bandwidth based on kernel function K(x) and the i+1th period Get the prediction type index of the i+1th period Corresponding prediction type exponential distribution The specific formula is as follows:

[0120]

[0121] Where N is the total number of nodes in a single cycle.

[0122] Furthermore, the distribution approximation unit obtains the predicted bus line loss rate distribution of the next cycle, including the following specific steps:

[0123] Call the bus line loss rate l of the i-th cycle from the cloud sharing module i As the predicted bus line loss rate distribution of the i+1th cycle The predicted distribution line loss mean

[0124] Call the forecast type exponential distribution for the i+1th period from the cloud sharing module And calculate the forecast type exponential distribution of the i+1th period The standard deviation of the forecast distribution index The specific calculation formula is as follows:

[0125]

[0126] in, is the exponential distribution of the prediction type in the i+1th period The exponential mean of the predicted distribution of

[0127] The standard deviation of the forecast distribution index As the predicted bus line loss rate distribution of the i+1th cycle The standard deviation of the predicted distribution line loss

[0128] Set the standard normal distribution to the mean of the predicted distribution line loss in the i+1th cycle and the standard deviation of the predicted distribution line loss Make corrections to obtain the predicted bus line loss rate distribution of the i+1th cycle The specific formula is as follows:

[0129]

[0130] Furthermore, the analysis decision unit adaptively determines the dynamic bus abnormality threshold of the next cycle, including the following specific steps:

[0131] Call the predicted bus line loss rate distribution of the i+1th cycle from the cloud sharing module Calculate the cumulative distribution of the predicted bus line loss rate for the i+1th cycle Predict the cumulative distribution of bus line loss rate Describes the predicted bus line loss rate in the i+1th cycle The probability of being less than or equal to a certain value, predicting the cumulative distribution of bus line loss rate The specific calculation formula is as follows:

[0132]

[0133] Further observation and prediction of the cumulative distribution of bus line loss rate Get the predicted bus line loss rate cumulative distribution The predicted bus line loss rate of the i+1th cycle corresponding to a value less than or equal to 0.8 is used as the dynamic bus abnormality threshold of the i+1th cycle

[0134] Further, the balance identification module includes an equation measurement unit and an identification decision unit;

[0135] The equation measurement unit calls the power sequence of the main transformer side and the branch line side of a single cycle at the end node of a single cycle, calculates the line loss error sequence of a single cycle in combination with the balance analysis model and uploads it to the cloud sharing module. The line loss error sequence includes the meter measurement error rate of all branch lines in a single cycle, the bus line loss rate and the bus fixed loss;

[0136] The identification decision unit calls the measurement error rate threshold ξ from the cloud sharing module β and a single-cycle line loss error sequence, dynamic bus anomaly threshold, and autonomously decide whether to send the branch line number and / or line loss error sequence to the maintenance correction module based on the progressive tracing strategy.

[0137] Furthermore, the equation measurement unit combines the balance analysis model to calculate the line loss error sequence of a single cycle, including the following specific steps:

[0138] Call the main transformer side power sequence of the i-th cycle from the cloud sharing module And the power sequence of the branch side in, is the nth node in the ith cycle The total power of the main transformer side, is the nth node in the ith cycle The charge vector on the branch line side, is the nth node in the ith cycle The amount of electricity on the mth branch line, where m is the branch line number, m = 1, ..., M, and M is the total number of branch lines;

[0139] The balance analysis equation is constructed to describe the balance relationship between the total power on the main transformer side and the total power on the branch line side of the bus at a single node. Then, the nth node in the ith cycle is The specific expression of the equilibrium analysis equation is as follows:

[0140]

[0141] Among them, l i and l0 are the bus line loss rate and bus fixed loss of the i-cycle that need to be calculated, respectively. is the measurement error rate of the branch line meter on the mth branch line on the branch line side of the i-th cycle to be calculated;

[0142] The equilibrium analysis equations of N nodes in the i-th period are combined and organized to construct the equilibrium analysis model. The specific expression of the equilibrium analysis model is as follows:

[0143]

[0144] Introduce the simplified matrix A of the i-th period i Solution vector x i Simplified equilibrium analysis model, the simplified form of the equilibrium analysis model is as follows:

[0145]

[0146] Among them, the simplified matrix A i Solution vector x i The specific forms are as follows:

[0147]

[0148] Based on linear algebra knowledge, we can solve the equilibrium analysis model and then solve the vector x i The specific formula is as follows

[0149]

[0150] Get the transpose of the solution vector (x i ) T , and transpose the solution vector (x i ) T The first M solutions are subtracted by 1 and then the inverse is taken to convert them into the line loss error sequence of the i-th cycle.

[0151] like Figure 3As shown, further, the identification decision unit autonomously decides whether to send the branch line number and / or line loss error sequence to the maintenance correction module based on the progressive tracing strategy, including the following specific steps:

[0152] Call the measurement error rate threshold ξ from the cloud sharing module β , the line loss error sequence of the i-th cycle and dynamic bus abnormality threshold

[0153] Determine the line loss error sequence of the i-th cycle The bus line loss rate l i Is it greater than the dynamic bus abnormality threshold of the i-th cycle?

[0154] If the bus line loss rate is l i Greater than the dynamic bus abnormality threshold This indicates that the bus is abnormal. It is further determined whether the metering error rate of the branch meter on each branch line in the i-th cycle is greater than the metering error rate threshold ξ β ;

[0155] If there is a measurement error rate of the mth branch line in the i-th cycle Greater than the measurement error rate threshold ξ β , indicating the measurement error rate Causes busbar abnormality, and the measurement error rate The corresponding branch line number m is sent to the maintenance and correction module, where m = 1, ..., M, and M is the total number of branch lines;

[0156] If the measurement error rate of the branch line meter on each branch line on the branch line side in the i-th cycle is less than or equal to the measurement error rate threshold ξ β , indicating that the bus abnormality comes from load imbalance, sending the line loss error sequence of the i-th cycle To the maintenance and correction module;

[0157] If the bus line loss rate is l i Less than the abnormal threshold of dynamic busbar This indicates that the busbar is normal. Further determine whether the metering error rate of the branch meter on each branch line in the i-th cycle is greater than the metering error rate threshold ξ β ;

[0158] If there is a measurement error rate of the mth branch line in the i-th cycle Greater than the measurement error rate threshold ξ β , indicating that there is a hidden danger in the mth branch line, and the measurement error rate The corresponding branch line number m is sent to the maintenance and correction module;

[0159] If the measurement error rate of the branch line meter on each line of the branch line side in the i-th cycle is less than or equal to the measurement error rate threshold ξ β , indicating that the busbar and all branch lines are normal and no operation is required.

[0160] Further, the maintenance correction module includes an electric meter maintenance unit, a coefficient calculation unit and an adjustment execution unit;

[0161] The maintenance correction module counts the branch line numbers received in a single cycle, integrates and generates branch line meter replacement prompts and sends them to the maintenance personnel's terminal;

[0162] The coefficient calculation unit obtains the line loss error sequence of a single cycle, generates the busbar regulation coefficient and branch line regulation coefficient vector of the next cycle based on the load regulation strategy, and sends them to the regulation execution unit;

[0163] The regulation execution unit includes a bus controller and M branch line controllers. The bus controller and the M branch line controllers are respectively installed in front of the bus gateway meter and the M branch line meters. The total power on the main transformer side and the power sequence on the branch line side can be adjusted at each node of the next cycle based on the bus regulation coefficient and the branch line regulation coefficient vector of the next cycle.

[0164] like Figure 4 As shown, further, the coefficient calculation unit generates the bus regulation coefficient and branch line regulation coefficient vector of the next cycle based on the load regulation strategy, including the following specific steps:

[0165] Get the line loss error sequence of the i-th cycle Call the main transformer side power sequence of the i-th cycle from the cloud sharing module Power sequence on the branch side and dynamic bus abnormality threshold

[0166] Establish the nth node of the i-th cycle The busbar regulation relationship is as follows:

[0167]

[0168] Among them, γ i+1 is the bus regulation coefficient of the i+1th cycle, is the branch line adjustment coefficient of the m-th branch line in the i+1-th period, m=1,…,M;

[0169] The bus regulation relationship of N nodes in the i-th period is combined and sorted to construct the bus regulation model. The specific expression is as follows:

[0170]

[0171] Call the maximum total power Q of the main transformer side from the cloud sharing module in,max , the minimum total power Q on the main transformer side in,min , Maximum power of branch line q out,max And the minimum power of the branch line q out,min , set the power regulation constraint group. Since the power of the bus and M branch lines cannot be scaled arbitrarily, the power regulation constraint group includes M+1 constraint inequalities, as follows:

[0172]

[0173] Due to the existence of the regulation power constraint group, the bus regulation model may not have a feasible solution. It is now rewritten as an optimization problem, and the right side of the bus regulation model equation is defined as the bus regulation coefficient γ for the i+1th cycle i+1 and M branch line adjustment coefficients The busbar regulation function The optimization objective is defined as minimizing the busbar regulation function, that is, Adjust the power constraint group as the constraint condition of the optimization problem;

[0174] The Lagrange multiplier method is used to introduce the constraints into the optimization objective, and the bus regulation coefficient γ of the i+1th cycle can be obtained by solving the KKT condition. i+1 and M branch line adjustment coefficients Combine M branch line adjustment coefficients Construct the line adjustment coefficient vector of the i+1th period

[0175] The present invention discloses a bus energy balance regulation system, comprising a system clock module, a data acquisition module, a three-phase balance module, a threshold adjustment module, a balance identification module, a maintenance correction module and a cloud sharing module. The system clock module sets nodes and cycles, and the data acquisition module collects data required by the system; the three-phase balance module adopts an intelligent reactive power compensation method to achieve dynamic three-phase balance; the threshold adjustment module considers the three-phase voltage imbalance and temperature, and determines the dynamic bus abnormality threshold based on the kernel density estimation method combined with the bus line loss rate; the balance identification module adopts a balance analysis model and a progressive tracing strategy to decide whether to send a branch line number and / or a line loss error sequence; the maintenance correction module generates a branch line meter replacement prompt based on the branch line number, and generates a bus adjustment coefficient and a branch line adjustment coefficient vector for the next cycle based on the load control strategy; the cloud sharing module realizes data sharing among various modules in the system.

[0176] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A busbar energy balance regulation system, characterized in that: It includes system clock module, data acquisition module, three-phase balance module, threshold adjustment module, balance identification module and maintenance correction module; The system clock module sets the node division period and announces it; The data acquisition module collects the data required by the system at each node; The three-phase balancing module adopts an intelligent reactive power compensation method to achieve dynamic three-phase balancing of a single node; The threshold adjustment module generates a type index sequence, uses a kernel density estimation method combined with a bus line loss rate to obtain a predicted bus line loss rate distribution, and determines a dynamic bus abnormality threshold; The balance identification module calculates and obtains the line loss error sequence based on the balance analysis model, and decides whether to send the branch line number and / or the line loss error sequence to the maintenance correction module based on the progressive tracing strategy; The maintenance correction module integrates and generates a reminder for replacing the branch line meter, and generates a busbar adjustment coefficient and a branch line adjustment coefficient vector for the next cycle based on the load regulation strategy and the line loss error sequence.

2. A busbar energy balance regulation system according to claim 1, characterized in that: The three-phase balancing module adopts the intelligent reactive power compensation method to realize the dynamic three-phase balancing of a single node, including the following specific steps: Obtaining the a-phase voltage, b-phase voltage, c-phase voltage, a-phase current, b-phase current and c-phase current of a single node in a single cycle, and the a-phase voltage, b-phase voltage and c-phase voltage can be mutually represented based on the phase offset; Based on the triangle load balancing formula, the load admittance of the ab phase, the load admittance of the bc phase and the load admittance of the ca phase are obtained, and the positive sequence current and the negative sequence current are further calculated to obtain the current imbalance degree; Obtain a current imbalance threshold and compare it with the current imbalance. When the current imbalance is greater than the current imbalance threshold, determine the compensated positive sequence current and the compensated negative sequence current based on the principle of symmetrical components. Based on the triangle compensation balance formula, the ab phase compensation admittance vector, the bc phase compensation admittance vector and the ca phase compensation admittance vector are derived.

3. A busbar energy balance regulation system according to claim 1, characterized in that: The threshold adjustment module includes an index prediction unit, a fluctuation fitting unit, a distribution approximation unit and an analysis and decision unit; The index prediction unit generates a type index sequence, and obtains a predicted type index sequence of the next period through LSTM; The fluctuation fitting unit adopts a kernel density estimation method to fit and generate a forecast type exponential distribution for the next period; The distribution approximation unit uses the prediction type exponential distribution of the next cycle to adjust the bus line loss rate of a single cycle to obtain the predicted bus line loss rate distribution of the next cycle; The analysis and decision-making unit determines a predicted dynamic bus abnormality threshold value for the next period based on the predicted bus line loss rate distribution for the next period.

4. A busbar energy balance regulation system as claimed in claim 3, characterized in that: The index prediction unit obtains the prediction type index sequence of the next period, including the following specific steps: Call and confirm the maximum phase voltage and the minimum phase voltage in the three-phase voltage vector of a single node in the three-phase voltage sequence of a single cycle, and calculate the voltage unbalance degree of the single node; Sort the voltage imbalance of all nodes in a single cycle to build a voltage imbalance sequence, call the temperature sequence of a single cycle and predict the bus line loss rate distribution; Resample the predicted bus line loss rate distribution at all nodes to obtain a single-cycle predicted bus line loss rate sequence; The voltage line loss correlation coefficient and the temperature line loss correlation coefficient of a single cycle are calculated respectively, and are used as the weights of the voltage imbalance degree sequence and the temperature sequence for weighted summation to generate a single cycle type index sequence; The type index sequence of a single cycle is input into LSTM to generate the predicted type index sequence of the next cycle.

5. A busbar energy balance regulation system as claimed in claim 3, characterized in that: The fluctuation fitting unit adopts the kernel density estimation method to fit and generate the forecast type exponential distribution of the next period, including the following specific steps: Call the prediction type index sequence of the next period and select the Gaussian function as the kernel function of the kernel density estimation; Determine the prediction bandwidth of the kernel function in the next cycle based on Silverman's rule of thumb; The predicted type exponential distribution of the next period is obtained based on the kernel function and the predicted bandwidth of the next period.

6. A busbar energy balance regulation system according to claim 1, characterized in that: The balance identification module includes an equation measurement unit and an identification decision unit; The equation measurement unit calls the power sequence of the main transformer side and the power sequence of the branch line side in a single cycle, calculates the line loss error sequence of a single cycle in combination with the balance analysis model and uploads it to the cloud sharing module; The identification decision unit independently decides whether to send the branch line number and / or line loss error sequence to the maintenance correction module based on the progressive tracing strategy.

7. A busbar energy balance regulation system according to claim 6, characterized in that: The equation measurement unit combines the balanced analysis model to calculate the line loss error sequence of a single cycle, including the following specific steps: Call the main transformer side power sequence and branch line side power sequence of a single cycle; The balance analysis equation is constructed to describe the balance relationship between the total power on the main transformer side and the total power on the branch line side at a single node in a single cycle of the busbar; The equilibrium analysis equations of all nodes in a single cycle are combined and organized to construct the equilibrium analysis model, and the simplified matrix and solution vector of a single cycle are further introduced for simplified representation; The balance analysis model is solved to obtain a solution vector, and the transpose of the solution vector is further transformed into a single-cycle line loss error sequence.

8. A busbar energy balance regulation system as claimed in claim 6, characterized in that: The identification decision unit autonomously decides whether to send the branch line number and / or line loss error sequence to the maintenance correction module based on the progressive tracing strategy, including the following specific steps: Calling the metering error rate threshold, the line loss error sequence of a single cycle and the dynamic bus abnormality threshold of a single cycle; Determine whether the bus line loss rate in the line loss error sequence of a single cycle is greater than the dynamic bus abnormality threshold of a single cycle; If it is greater than the abnormal threshold of the dynamic busbar in a single cycle, further determine whether the metering error rate of each branch line in a single cycle is greater than the metering error rate threshold; If the measurement error rate of a certain branch line in a single cycle is greater than the measurement error rate threshold, the corresponding branch line number is sent to the maintenance correction module; If the metering error rates of all branch lines in a single cycle are less than or equal to the metering error rate threshold, the line loss error sequence of a single cycle is sent to the maintenance correction module; If it is less than the abnormal threshold of the dynamic busbar in a single cycle, further determine whether the metering error rate of each branch line in a single cycle is greater than the metering error rate threshold; If the metering error rate of a certain branch line in a single cycle is greater than the metering error rate threshold, the corresponding branch line number is sent to the maintenance correction module.

9. A busbar energy balance regulation system according to claim 1, characterized in that: The maintenance correction module includes an electric meter maintenance unit, a coefficient calculation unit and an adjustment execution unit; The maintenance correction module integrates the branch line numbers received in a single cycle to generate branch line meter replacement prompts; The coefficient calculation unit obtains the line loss error sequence of a single cycle, generates the busbar regulation coefficient and branch line regulation coefficient vector of the next cycle based on the load regulation strategy, and sends them to the regulation execution unit; The regulation execution unit performs regulation at each node in the next cycle based on the bus regulation coefficient and the branch regulation coefficient vector.

10. A busbar energy balance regulation system according to claim 9, characterized in that: The coefficient calculation unit generates the busbar regulation coefficient and branch line regulation coefficient vector for the next cycle based on the load regulation strategy, including the following specific steps: Obtain the line loss error sequence of a single cycle, call the main transformer side power sequence, branch line side power sequence, main transformer side maximum total power, main transformer side minimum total power, branch line maximum power and branch line minimum power of a single cycle; Establish a bus regulation relationship for a single node in a single cycle, combine the bus regulation relationships of all nodes in a single cycle to generate a bus regulation model, and set a regulation power constraint group; The right side of the bus regulation model equation is defined as the bus regulation function, the optimization objective is defined as minimizing the bus regulation function, and the regulation power constraint group is used as the constraint condition of the optimization problem; The Lagrange multiplier method is used to introduce constraints into the optimization objective. The busbar adjustment coefficient and all branch line adjustment coefficients of the next period are obtained by solving the KKT conditions, and all branch line adjustment coefficients are combined to construct the branch line adjustment coefficient vector.

Images