Power grid cable loss analysis method and device and storage medium

Abstract

The invention provides a power grid cable loss analysis method and device and a storage medium, and the method comprises the steps: obtaining cable topology data, extracting parameter characteristics, and generating a parameter database; performing matrix conversion based on the parameter database to establish a network calculation model, executing load flow calculation, and outputting voltage power distribution data; establishing a temperature and loss coupling relation model according to the voltage power distribution data, calculating cable temperature distribution, and generating time-varying loss data; and performing multi-dimensional feature analysis on the time-varying loss data, calculating a loss risk index, identifying loss hotspots and generating an optimization scheme. According to the method, the dynamic relation model of the cable parameters and the loss is established, and the cable shielding layer characteristics and the temperature change factors are combined, so that accurate analysis of the power grid cable loss is realized, technical support is provided for identifying loss hotspots and formulating a targeted optimization scheme, and the power grid loss calculation deviation is effectively reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of cable loss analysis, and in particular, to a method, device, and storage medium for analyzing power grid cable losses. Background Art

[0002] With the continuous expansion of the scale of the power system and the continuous increase of the load, the problem of power grid cable losses has become increasingly prominent. Cable losses not only directly affect the economic operation of the power grid, but also shorten the service life of the cable and increase the risk of system failures. The existing methods for analyzing power grid cable losses mainly calculate based on static parameter models, and usually directly estimate the losses by using the square relationship between the rated parameters of the cable and the load current.

[0003] However, there is a prominent problem in the existing technology: it cannot accurately reflect the dynamic relationship between cable characteristics and losses. The traditional method ignores the influence of cable temperature changes on the resistance value, does not consider the effect of the grounding method of the cable shielding layer on the losses, and does not fully pay attention to key factors such as the cable laying environment and joint positions, resulting in a large deviation between the loss calculation result and the actual situation. Especially in urban distribution networks with large load fluctuations and complex laying environments, it seriously affects the accuracy of power grid planning and renovation decisions. Summary of the Invention

[0004] The main purpose of the present invention is to solve the technical problem that the power grid cable loss analysis method cannot accurately reflect the dynamic relationship between cable characteristics and losses; In the first aspect of the present invention, a method for analyzing power grid cable losses is provided, and the method for analyzing power grid cable losses includes: Obtain the topological data of the cables in the power grid system, extract the cable parameter characteristics according to the topological data, and generate a cable parameter database; Based on the cable parameter database, perform parameter matrix conversion on the cable network in the power grid system and establish a network calculation model, execute power flow calculation through the network calculation model, and output the voltage power distribution data of the cable network; According to the voltage power distribution data, establish a coupling relationship model between cable temperature and losses, calculate the temperature distribution of the cable under different conditions through the coupling relationship model, and generate the time-varying loss data of the cable network; Perform multi-dimensional feature analysis on the time-varying loss data, calculate the cable loss risk index through the multi-dimensional feature analysis, identify the loss hotspots according to the loss risk index, and generate a cable optimization plan.

[0005] Optionally, in the first implementation manner of the first aspect of the present invention, the obtaining the topological data of the cables in the power grid system, extracting the cable parameter characteristics according to the topological data, and generating a cable parameter database includes: Collect the connection relationship data between cables through the automation equipment of the power system to form topology data; Analyze the grounding method of the cable shielding layer according to the topology data, and calculate the corresponding shielding layer current path coefficient for various grounding methods; Based on the cable connection relationship data and the service life records of the cables, allocate aging degradation parameters to each cable; Utilize the shielding layer current path coefficient and the aging degradation parameters, and calculate the equivalent impedance values corresponding to each parameter combination through electromagnetic field numerical simulation to generate a cable parameter database.

[0006] Optionally, in the second implementation manner of the first aspect of the present invention, the performing parameter matrix conversion on the cable network in the power grid system and establishing a network calculation model based on the cable parameter database, and performing power flow calculation through the network calculation model, and outputting the voltage power distribution data of the cable network includes: Construct a node-branch incidence matrix of the cable network according to the cable parameter database, and divide the cable network into two types of ports, namely load nodes and power supply nodes, based on the node-branch incidence matrix; Construct a hybrid parameter matrix based on the two types of ports and the equivalent impedance values in the cable parameter database, and convert the hybrid parameter matrix into an admittance matrix through matrix operations; Calculate the correction factor of the shielding layer circulating current according to the cable shielding layer information in the cable parameter database, apply the correction factor to the admittance matrix, and generate a corrected admittance matrix, and the corrected admittance matrix constitutes the network calculation model; Utilize the corrected admittance matrix in the network calculation model to perform power flow calculation by applying the improved Newton-Raphson iteration algorithm to obtain the voltage power distribution data of the cable network.

[0007] Optionally, in the third implementation manner of the first aspect of the present invention, the performing power flow calculation by applying the improved Newton-Raphson iteration algorithm using the corrected admittance matrix in the network calculation model to obtain the voltage power distribution data of the cable network includes: Construct a node voltage balance equation set for power flow calculation according to the corrected admittance matrix, and set an initial voltage value as the iteration starting point for the node voltage balance equation set; Calculate the Jacobian matrix for the node voltage balance equation set, and solve the power flow node voltage correction amount by using the improved Newton-Raphson iteration method based on the Jacobian matrix; Update the node voltage value in the power flow calculation according to the node voltage correction amount, and judge the power flow convergence of the node voltage value. If the power flow convergence condition is not reached, return to the previous step and continue the iterative calculation; Using the node voltage values after power flow convergence, calculate the power flow current distribution of each branch in the cable network according to Ohm's law; According to the node voltage values and the power flow current distribution, calculate the power flow power distribution of each section of the cable in the cable network to obtain the voltage-power distribution data of the cable network.

[0008] Optionally, in the fourth implementation manner of the first aspect of the present invention, the establishing a coupling relationship model between cable temperature and loss according to the voltage-power distribution data, and calculating the temperature distribution of the cable under different conditions through the coupling relationship model to generate the time-varying loss data of the cable network includes: Calculate the current value and initial loss value of each section of the cable according to the voltage-power distribution data; Based on the current value and the initial loss value, construct a coupling relationship model between cable temperature and loss that mutually affects current heating, temperature rise, and resistance change. The coupling relationship model is characterized by a cable load-loss-temperature dynamic response equation; According to the laying method information and ambient temperature data in the topology data, determine the heat dissipation parameters of each section of the cable, and integrate the heat dissipation parameters into the coupling relationship model; Using the coupling relationship model integrated with heat dissipation parameters, adopt a time-step rolling calculation method to iteratively solve the temperature distribution of the cable under a typical load curve; According to the calculated temperature distribution and the resistance-temperature coefficient relationship in the coupling relationship model, update the actual resistance value of the cable in each period, and calculate and output the time-varying loss data.

[0009] Optionally, in the fifth implementation manner of the first aspect of the present invention, the updating the actual resistance value of the cable in each period according to the calculated temperature distribution and the resistance-temperature coefficient relationship in the coupling relationship model, and calculating and outputting the time-varying loss data includes: According to the calculated temperature distribution, calculate the actual resistance value of each section of the cable in the cable network for each period; Substitute the actual resistance value into Ohm's law, and combine the current value in the voltage-power distribution data to calculate the line loss data of the cable in each period; Identify the cable joint positions according to the laying method information, calculate the additional loss at the joint positions, and incorporate the additional loss into the line loss data; Perform time series integration on the line loss data in each period to obtain a time-varying loss data matrix, and calculate the average loss level, loss time-varying characteristics, and loss spatial distribution characteristics of each cable section based on the time-varying loss data matrix, and output the time-varying loss data of the cable network.

[0010] Optionally, in the sixth implementation manner of the first aspect of the present invention, the multi-dimensional feature analysis of the time-varying loss data, calculating the cable loss risk index through the multi-dimensional feature analysis, identifying the loss hotspots according to the loss risk index, and generating the cable optimization plan include: Extracting the characteristic parameters of three dimensions, namely the time dimension, the space dimension, and the operating condition dimension, from the time-varying loss data respectively; Calculating the cable loss risk index of each section of the cable through a feature fusion algorithm according to the characteristic parameters of the three dimensions; Based on the cable loss risk index and a preset risk threshold, identifying the cable sections with loss risks exceeding the threshold as loss hotspots, analyzing the distribution of the characteristic parameters of the loss hotspots, determining the dominant characteristic parameter combination of each loss hotspot, and defining the dominant characteristic parameter combination as the characteristic mode of the loss hotspot; For the loss hotspots with different characteristic modes, according to the topological data, generating corresponding cable optimization plans through a multi-level cable optimization decision tree.

[0011] The second aspect of the present invention provides a power grid cable loss analysis device, and the power grid cable loss analysis device includes: A data acquisition module, configured to obtain the topological data of the cables in the power grid system, extract the cable parameter characteristics according to the topological data, and generate a cable parameter database; A power flow calculation module, configured to perform parameter matrix conversion on the cable network in the power grid system based on the cable parameter database, establish a network calculation model, perform power flow calculation through the network calculation model, and output the voltage power distribution data of the cable network; A loss calculation module, configured to establish a coupling relationship model between the cable temperature and the loss according to the voltage power distribution data, calculate the temperature distribution of the cable under different conditions through the coupling relationship model, and generate the time-varying loss data of the cable network; A hotspot identification module, configured to perform multi-dimensional feature analysis on the time-varying loss data, calculate the cable loss risk index through the multi-dimensional feature analysis, identify the loss hotspots according to the loss risk index, and generate a cable optimization plan.

[0012] The third aspect of the present invention provides a computer-readable storage medium, in which instructions are stored, and when it runs on a computer, it causes the computer to execute the steps of the above-mentioned power grid cable loss analysis method.

[0013] The above power grid cable loss analysis method, device and storage medium obtain cable topology data, extract parameter characteristics, and generate a parameter database; based on the parameter database, perform matrix conversion to establish a network calculation model, execute power flow calculation, and output voltage and power distribution data; establish a coupling relationship model between temperature and loss according to the voltage and power distribution data, calculate the cable temperature distribution, and generate time-varying loss data; perform multi-dimensional feature analysis on the time-varying loss data, calculate the loss risk index, identify loss hotspots and generate optimization solutions. By establishing a dynamic relationship model between cable parameters and losses and combining the characteristics of the cable shielding layer and temperature change factors, the present invention realizes accurate analysis of power grid cable losses, provides technical support for identifying loss hotspots and formulating targeted optimization solutions, and effectively reduces the calculation deviation of power grid losses.

[0014] Other features and advantages of the present invention will be described in the following specification, and, in part, will be obvious from the specification, or will be understood by implementing the present invention. The objectives and other advantages of the present invention are realized and obtained by the structures specifically pointed out in the specification, claims and drawings.

[0015] To make the above objectives, features and advantages of the present invention more obvious and understandable, the following specific preferred embodiments are given and, in conjunction with the accompanying drawings, are described in detail as follows. Brief Description of the Drawings

[0016] Figure 1 It is a schematic diagram of an embodiment of the power grid cable loss analysis method in an embodiment of the present invention; Figure 2 It is a schematic diagram of an embodiment of the power grid cable loss analysis device in an embodiment of the present invention. Detailed Description of the Embodiment

[0017] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0018] The terms "including" and "having" and any variations thereof mentioned in the embodiments of the present invention are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device including a series of steps or units is not limited to the listed steps or units, but optionally further includes other unlisted steps or units, or optionally further includes other steps or units inherent to these processes, methods, products or devices.

[0019] To facilitate the understanding of this embodiment, a method for analyzing the losses of power grid cables disclosed in the embodiments of the present invention will be introduced in detail first. As Figure 1 shown, this method includes the following steps: 101. Obtain the topological data of the cables in the power grid system, extract the cable parameter characteristics according to the topological data, and generate a cable parameter database; In an embodiment of the present invention, the obtaining the topological data of the cables in the power grid system, extracting the cable parameter characteristics according to the topological data, and generating a cable parameter database includes: collecting the connection relationship data between the cables through the automation equipment of the power system to form topological data; analyzing the grounding method of the cable shielding layer according to the topological data, and calculating the corresponding shielding layer current path coefficient for various grounding methods; allocating aging degradation parameters to each cable based on the cable connection relationship data and the service life records of the cables; using the shielding layer current path coefficient and the aging degradation parameters, calculating the equivalent impedance values corresponding to each parameter combination through electromagnetic field numerical simulation, and generating a cable parameter database.

[0020] Specifically, it is first necessary to obtain the topological data of the cables in the power grid system, extract the cable parameter characteristics according to the topological data, and generate a cable parameter database. This process collects the connection relationship data between the cables through the automation equipment of the power system to form topological data. In this step, equipment such as a distribution automation system (DAS), a substation automation system (SAS), and an intelligent electronic device (IED) are used to collect data such as the physical connection status of the cables, node position information, cable model specifications, and laying methods. For example, the geographic information module in the distribution management system is used to obtain the spatial positioning information of the cables, including the start and end point coordinates and the position data of the intermediate connection nodes; at the same time, the cable specification information, such as basic parameters such as cross-sectional size, insulation type, and conductor material, is extracted from the equipment asset management system. After these data are uniformly formatted and cleaned, a topological data set containing node information and connection relationships is constructed, and this data set contains key information such as the line numbers, start and end nodes, lengths, and installation dates of the cables.

[0021] Specifically, based on the collected topology data, the grounding method of the cable shielding layer is further analyzed, and the corresponding shielding layer current path coefficients are calculated for various grounding methods. The grounding methods of the cable shielding layer usually include types such as single-point grounding, two-end grounding, cross-bonding, and multi-point grounding at special positions. For the single-point grounding method, the shielding layer is connected to the ground at only one end, and there is no longitudinal current in the shielding layer, so the shielding layer current path coefficient is relatively small; for the two-end grounding method, the shielding layer is connected to the ground at both ends, forming a complete loop, resulting in an induced circulating current, and its shielding layer current path coefficient is relatively large; while for the cross-bonding grounding method, by exchanging the connection positions of the three-phase shielding layers at regular intervals (usually 1 / 3 of the cable section length), the induced electromotive force of the shielding layer can be effectively canceled, reducing the shielding layer current, and at this time the current path coefficient is at a medium level. By analyzing the grounding methods of the shielding layers of each section of the cable in the power grid and combining factors such as the cable length, cross-sectional area, and installation status, a calculation model for the shielding layer current path coefficient is established. This model takes into account the geometric dimensions, material characteristics, and electromagnetic field distribution of the shielding layer, and finally obtains the path coefficient values that quantify the influence degrees of various grounding methods.

[0022] Specifically, based on the cable connection relationship data and the service life records of the cables, appropriate aging degradation parameters are assigned to each cable. During long-term operation, cables will experience performance degradation due to factors such as thermal aging, oxidation, humidity, mechanical stress, and electrical stress. This degradation directly affects the electrical parameters of the cables. According to the commissioning date of the cables recorded in the power equipment management system and combined with the typical life curve of this type of cable, the aging degree of each section of the cable is calculated. At the same time, considering the operating environment factors of the cables, such as the average temperature, humidity, groundwater level, and load history of the laying area, the aging degradation rate is corrected. For example, for a 10kV cross-linked polyethylene insulated power cable that has been in operation for 15 years, under standard environmental conditions, its insulation resistance will decrease by about 15%, and the conductor resistance will increase by about 3% due to oxidation; while if this cable is located in a high-temperature and humid environment, the degradation degree will be more serious. By establishing an aging degradation model that comprehensively considers time factors and environmental factors, specific aging degradation parameters are assigned to each section of the cable in the power grid, and these parameters directly reflect the current actual electrical performance status of the cables.

[0023] Specifically, using the shielding layer current path coefficient and aging degradation parameters calculated above, the equivalent impedance values corresponding to each parameter combination are calculated through electromagnetic field numerical simulation technology, and finally a cable parameter database is generated. In this step, a three-dimensional electromagnetic field model of the cable is constructed using the finite element method or the boundary element method to simulate the electric and magnetic field distributions inside the cable under different working conditions. Electromagnetic field simulation needs to consider the cable's geometric structure (such as conductor diameter, insulation thickness, shielding layer thickness, etc.), material properties (such as conductor conductivity, insulation dielectric constant, etc.), and laying environment (such as soil resistivity, influence of surrounding cables, etc.). During the simulation, the previously obtained shielding layer current path coefficient and aging degradation parameters are used as key inputs to adjust the corresponding model parameters, such as conductor resistivity and shielding layer conductivity. By solving Maxwell's equations, the distribution parameters of the cable at different frequencies are calculated, including longitudinal impedance and transverse admittance. These distribution parameters are further converted into equivalent resistance, inductance, and capacitance values in the lumped parameter circuit model to form a complete cable parameter database. This database contains the actual electrical parameters of each section of the cable in the power grid under different operating conditions, accurately reflecting the current physical state of the cable.

[0024] 102. Based on the cable parameter database, perform parameter matrix conversion on the cable network in the power grid system and establish a network calculation model. Execute power flow calculation through the network calculation model and output the voltage-power distribution data of the cable network; In an embodiment of the present invention, the step of performing parameter matrix conversion on the cable network in the power grid system based on the cable parameter database, establishing a network calculation model, and executing power flow calculation through the network calculation model to output the voltage-power distribution data of the cable network includes: constructing a node-branch incidence matrix of the cable network according to the cable parameter database, and dividing the cable network into two types of ports, namely load nodes and power supply nodes, based on the node-branch incidence matrix; constructing a hybrid parameter matrix based on the two types of ports and the equivalent impedance values in the cable parameter database, and converting the hybrid parameter matrix into an admittance matrix through matrix operations; calculating a correction factor for the shielding layer circulating current according to the cable shielding layer information in the cable parameter database, applying the correction factor to the admittance matrix to generate a corrected admittance matrix, and the corrected admittance matrix constitutes the network calculation model; using the corrected admittance matrix in the network calculation model to apply the improved Newton-Raphson iteration algorithm to execute power flow calculation to obtain the voltage-power distribution data of the cable network.

[0025] Specifically, after obtaining the cable parameter database, it is necessary to perform parameter matrix conversion on the cable network in the power grid system based on the cable parameter database and establish a network calculation model, and perform power flow calculation through the network calculation model to output the voltage and power distribution data of the cable network. First, construct a node-branch incidence matrix of the cable network according to the cable parameter database, and divide the cable network into two types of ports, namely load nodes and power supply nodes, based on the node-branch incidence matrix. Specifically, the node-branch incidence matrix A is the basic matrix describing the topology of the power grid, and this matrix represents the connection relationship between each node and each branch in the power grid. During the construction process, first number all the nodes and branches in the cable network. For example, the outgoing line nodes of the substation, the connection points of the distribution transformers, the ring main unit nodes, etc. are set as network nodes, and the cable segments connecting each node are used as network branches. Then, fill in the element values of the node-branch incidence matrix according to the cable connection relationship: when the starting end of branch j is node i, the matrix element A(i,j) takes the value of 1; when the terminal end of branch j is node i, the matrix element A(i,j) takes the value of -1; in other cases, it takes the value of 0. For example, in a certain 10 kV distribution network, there is a cable numbered L105 connecting the outgoing line node N023 of the substation and the ring main unit node N047, then 1 and -1 are filled in the corresponding positions in the incidence matrix respectively. After completing the construction of the node-branch incidence matrix, divide the entire cable network into two types of ports according to the electrical characteristics of the nodes: load nodes and power supply nodes. Load nodes include various power consumption connection points, such as the primary sides of distribution transformers in industrial users, commercial areas, and residential communities. These nodes usually consume electrical energy; while power supply nodes include busbars of substations at all levels, connection points of power generation equipment, etc. These nodes usually provide electrical energy. This division clarifies the direction of energy flow in the cable network and lays a foundation for the subsequent construction of the hybrid parameter matrix.

[0026] Specifically, based on the two types of ports divided above and the equivalent impedance values in the cable parameter database, a hybrid parameter matrix is constructed, and the hybrid parameter matrix is converted into an admittance matrix through matrix operations. The hybrid parameter matrix H is a multi-port network parameter representation form that can handle voltage sources and current sources simultaneously, and is particularly suitable for describing the complex relationship between load nodes and power supply nodes. When constructing the hybrid parameter matrix, first, the voltage and current of the load node are used as type I port variables, and the voltage and current of the power supply node are used as type II port variables. The four sub-blocks of the hybrid parameter matrix respectively represent: H11 represents the relationship between the voltage and current of the load node, H12 represents the relationship between the voltage of the load node and the voltage of the power supply node, H21 represents the relationship between the current of the power supply node and the current of the load node, and H22 represents the relationship between the current of the power supply node and the voltage. When filling the element values of the hybrid parameter matrix, using the equivalent impedance data in the cable parameter database, combined with the node-branch incidence matrix, the specific values of each sub-block are calculated through the principle of network analysis. For example, in a certain urban distribution network, a YJV22-3×120 cable that has been used for 15 years connects the substation and the distribution room. The equivalent impedance recorded in the cable parameter database is 0.2 + j0.18 Ω / km, and the length is 1.5 km. Then, the values of the corresponding elements in the hybrid parameter matrix are calculated using this parameter. After completing the construction of the hybrid parameter matrix, it is necessary to convert it into an admittance matrix form that is more suitable for power flow calculation. During the conversion process, first, the inverse matrix of the H11 sub-block is solved, and then, using the matrix block operation rule, the hybrid parameter matrix H is converted into an admittance matrix Y that represents the relationship between the node voltage and the injected current. The diagonal elements of the admittance matrix Y represent the sum of the admittances of all branches connected to each node, and the non-diagonal elements represent the negative values of the admittances of the branches connecting the nodes.

[0027] Specifically, a correction factor for the shield current of the cable is calculated based on the cable shield information in the cable parameter database, and the correction factor is applied to the admittance matrix to generate a corrected admittance matrix, which constitutes the network calculation model. The shield current of the cable has a significant impact on the effective impedance of the cable. The traditional admittance matrix does not consider this factor, resulting in calculation deviations. The shield current is mainly caused by three-phase unbalanced current, harmonic current, and differences in grounding methods, and its magnitude is closely related to the cable shield resistance, induction coupling degree, and grounding loop impedance. When calculating the correction factor of the shield current, first extract information such as the shield resistance, mutual inductance between the shield and the conductor, and the shield grounding method from the cable parameter database. For a single-point grounded cable, since there is no circulating current, the correction factor is close to 1; for a two-end grounded cable, the additional impedance caused by the induced current is calculated according to the electromagnetic coupling degree between the shield and the conductor, and its correction factor is usually greater than 1; for a cross-bonded grounded cable, based on the relationship between the cross-bonding spacing and the cable section length, the partially cancelled circulating current effect is calculated to obtain a correction factor between the previous two. For example, a 35 kV power cable laid in an urban underground cable tunnel adopts a two-end grounding method, and the copper cross-section of its shield is 25 mm². The correction factor of the shield current calculated by electromagnetic analysis is 1.12, indicating that the actual effective impedance is 12% higher than the theoretical value. Apply the calculated correction factor to the corresponding elements of the admittance matrix. According to the relationship Yij' = Yij / αij (where αij is the correction factor of the shield current of the cable section connecting nodes i and j), correct the values of each element in the admittance matrix to obtain the corrected admittance matrix Y'. This corrected admittance matrix more accurately reflects the actual electrical characteristics of the cable network and constitutes a complete network calculation model.

[0028] Specifically, the modified admittance matrix in the constructed network calculation model is utilized, and the improved Newton-Raphson iterative algorithm is applied to perform power flow calculation to obtain the voltage-power distribution data of the cable network. The Newton-Raphson algorithm is a classical numerical method for solving non-linear equations and is widely used in power system power flow calculation. The improved Newton-Raphson algorithm is optimized based on the standard algorithm to solve the convergence problem caused by the low resistance and high susceptance characteristics of the cable network. The core of power flow calculation is to solve the nodal power balance equation, that is, the injected power at each node is equal to the sum of the powers of all connecting branches passing through this node. In the calculation process, the initial voltage values of each node are first set. Usually, the power supply node is set to the rated voltage (for example, the 10 kV distribution network is set to 10.5 kV), and the load node is set to 1.0 per unit value. Then, based on the current voltage values, the nodal power imbalance is calculated using the modified admittance matrix, the Jacobian matrix is constructed, and the correction amount of the nodal voltage is solved. In the cable network, the influence of the shield layer circulating current is reflected in the modified admittance matrix, making the calculated power imbalance more accurate. After each iteration, the convergence condition is checked. Usually, setting the maximum power imbalance less than a predetermined threshold (such as 0.0001 p.u.) is used as the convergence criterion. For complex loop network structures or heavily loaded cable networks, the standard Newton-Raphson algorithm may face convergence difficulties. The improved algorithm adopts techniques such as adaptive step size adjustment and imbalance weighting processing to improve convergence. For example, in a loop distribution system in the core area of a city, which contains 47 nodes and 56 cable branches, the standard algorithm requires 12 iterations and there is a risk of oscillation, while the improved algorithm only needs 8 iterations to converge stably. After the power flow calculation converges, the voltage amplitude and phase angle of each node in the cable network are obtained, and then the current, active power, and reactive power of each cable branch are calculated. These data together constitute the voltage-power distribution data, comprehensively describing the operating state of the cable network.

[0029] Further, the performing power flow calculation using the modified admittance matrix in the network calculation model and obtaining the voltage-power distribution data of the cable network by the improved Newton-Raphson iterative algorithm includes: constructing a nodal voltage balance equation set for power flow calculation according to the modified admittance matrix, and setting the initial voltage value as the iteration starting point for the nodal voltage balance equation set; calculating the Jacobian matrix for the nodal voltage balance equation set, and solving the correction amount of the power flow nodal voltage by using the improved Newton-Raphson iterative method based on the Jacobian matrix; updating the nodal voltage value in the power flow calculation according to the correction amount of the nodal voltage, and judging the power flow convergence of the nodal voltage value. If the power flow convergence condition is not reached, return to the previous step and continue the iterative calculation; using the nodal voltage value after power flow convergence, calculating the power flow current distribution of each branch in the cable network according to Ohm's law; calculating the power flow power distribution of each section of the cable in the cable network according to the nodal voltage value and the power flow current distribution, and obtaining the voltage-power distribution data of the cable network.

[0030] Specifically, in the analysis of power grid cable losses, performing power flow calculations through the improved Newton-Raphson iterative algorithm using the modified admittance matrix in the network calculation model to obtain the voltage-power distribution data of the cable network is a multi-step and delicate process. First, based on the modified admittance matrix, a nodal voltage balance equation set for power flow calculations is constructed, and an initial voltage value is set for these equations as the starting point of the iteration. In the specific implementation process, the nodal voltage balance equation set reflects the basic physical law of the power system - Kirchhoff's current law, and this equation set expresses that the net power injected into each node is equal to the sum of the power transmitted through all connected branches at that node. For any node i in the cable network, its injected power can be expressed as , where is the active power, is the reactive power. Using the element values of the modified admittance matrix , after establishing the voltage balance equation set, appropriate initial voltage values for each node are set. Usually, for PV nodes (such as generator connection points), they are set to the rated voltage value and the specified power value, and for PQ nodes (such as load points), the initial voltage value is set to 1.0 p.u. 20°. These initial values constitute the starting point of the iterative calculation. Next, the Jacobian matrix is calculated, and based on the Jacobian matrix, the improved Newton-Raphson iterative method is used to solve for the correction amount of the nodal voltage. The Jacobian matrix represents the partial derivatives of the nodal power with respect to the nodal voltage magnitude and phase angle, and its structure includes four submatrices: . In the cable network, the correction factor of the shield layer circulating current indirectly affects the element values of the Jacobian matrix through the modified admittance matrix. The traditional Jacobian matrix calculation method has numerical stability problems in the cable network, so an enhanced partial derivative calculation method is adopted: ; where, is the active power of node , is the voltage phase angle of node , and are the voltage magnitudes of nodes and , and are the real and imaginary parts in the modified admittance matrix, is the voltage phase angle difference between nodes and , is the correction function considering the polarization effect of the cable insulation medium, is the tangent of the dielectric loss angle of the cable between nodes and , To consider the node the correction function of the equivalent grounding resistance of the shielding layer at For the node the equivalent grounding resistance of the shielding layer at is the total number of network nodes. Update the node voltage value in the power flow calculation according to the calculated node voltage correction amount, and judge whether these voltage values meet the power flow convergence condition. If the convergence condition is not met, return to the previous iteration step to continue the calculation. The updated node voltage is calculated by adding the current value and the correction amount , where represents the number of iterations. After the voltage is updated, it is necessary to recalculate the power imbalance of each node to evaluate the convergence degree of the current solution. For the cable network, a multi-criterion convergence determination method is adopted, which not only examines the power imbalance, but also monitors the change trend of the voltage correction amount.

[0031] Using the node voltage value after power flow convergence, calculate the power flow current distribution of each branch in the cable network according to Ohm's law. The current calculation of the cable branch needs to consider the relationship between the voltage difference at both ends of the node and the branch impedance. For the cable branch connecting node i and j, its complex current value can be calculated by the following formula: Among them, is the complex value of the branch current from the node to the node , is the corrected branch admittance, and are the complex voltages of the nodes and , is the branch-to-ground admittance, is the cable characteristic admittance correction function, is the system frequency, is the cable laying method parameter, is the cable burial depth. According to the calculated node voltage value and branch power flow current distribution, calculate the power flow power distribution of each section of the cable in the cable network, and finally obtain the voltage power distribution data of the cable network. The complex power calculation of the cable branch needs to consider the relationship between the current flow direction and the voltage reference direction. For the complex power flowing from the node to the node can be expressed as: Among them is the conjugate complex number of the current , and is further decomposed into active power and reactive power: Since the cable has obvious distributed parameter characteristics, the power calculation of a long cable also needs to consider the distributed capacitance loss along the line, which is often ignored in traditional calculations. By calculating the power flow direction and loss distribution of each section of the cable, a complete voltage-power distribution data set is finally formed, which includes the voltage amplitude and phase angle of all nodes in the network, the current amplitude and phase angle of all branches, and the active power, reactive power and loss data of each branch.

[0032] 103. According to the voltage-power distribution data, establish a coupling relationship model between cable temperature and loss, and calculate the temperature distribution of the cable under different conditions through the coupling relationship model to generate time-varying loss data of the cable network; In an embodiment of the present invention, the establishing a coupling relationship model between cable temperature and loss according to the voltage-power distribution data, and calculating the temperature distribution of the cable under different conditions through the coupling relationship model to generate time-varying loss data of the cable network includes: calculating the current value and initial loss value of each section of the cable according to the voltage-power distribution data; based on the current value and initial loss value, constructing a coupling relationship model between cable temperature and loss that includes the mutual influence of current heating, temperature rise and resistance change, and the coupling relationship model is characterized by a cable load-loss-temperature dynamic response equation; determining the heat dissipation parameters of each section of the cable according to the laying method information and environmental temperature data in the topology data, and integrating the heat dissipation parameters into the coupling relationship model; using the coupling relationship model integrated with heat dissipation parameters, adopting a time-step rolling calculation method to iteratively solve the temperature distribution of the cable under a typical load curve; according to the calculated temperature distribution and the resistance-temperature coefficient relationship in the coupling relationship model, updating the actual resistance value of the cable in each time period, and calculating and outputting time-varying loss data.

[0033] Specifically, based on the voltage-power distribution data, a coupling relationship model between cable temperature and loss is established. The implementation process of calculating the temperature distribution of the cable under different conditions through the coupling relationship model and generating the time-varying loss data of the cable network first requires calculating the current value and initial loss value of each section of the cable according to the voltage-power distribution data. In specific operations, using the voltage-power distribution data obtained from power flow calculations, the node voltages and current values at both ends of each section of the cable are extracted. The calculation of the current value is based on Ohm's law, which is obtained by dividing the voltage difference between the two end nodes by the cable impedance. For example, for a 10kV YJV22-3×120 cable connecting the outgoing cabinet of a substation and a distribution box, with a length of 0.8 km, through the voltage-power distribution data, the voltage at the head end is 10.2 kV and the voltage at the tail end is 10.15 kV, then the current value is approximately 98.7 A. The calculation of the initial loss value uses the traditional I²R method, that is, the square of the current multiplied by the cable resistance. The resistance value used here is taken from the initial value in the cable parameter database, which has not considered the influence of temperature and is the result under ideal calculation conditions. For the above cable section, if the resistance at 20°C is 0.153 Ω / km, then the calculated value of the initial loss is approximately 1.19 kW. It should be noted that this initial loss value is only used as the starting point for the coupling calculation and is different from the loss at the actual operating temperature of the cable. In the actual power grid, the distribution of the initial loss values of the cables is uneven. Usually, the initial loss values are higher near substations, in load center areas, and at cable cross-laying locations, and these areas are often the locations where the temperature rises most significantly. By calculating the current values and initial loss values of each section of the entire cable network, a complete initial state of the loss distribution is formed, and this state will be used as the basic data for the subsequent coupling calculation of temperature and loss.

[0034] Specifically, based on the calculated current value and the initial loss value, a coupling relationship model between cable temperature and loss that includes the mutual influence of current heating, temperature rise, and resistance change is constructed. The coupling relationship model is characterized by a cable load-loss-temperature dynamic response equation. There is a positive feedback mechanism between cable temperature and loss: the current flowing through the cable generates loss heat, resulting in a temperature rise, and the temperature rise further increases the cable resistance, further increasing the loss, forming a cyclic influence. The core of this coupling relationship model lies in establishing an accurate cable thermodynamics model and a relationship model of electrical characteristics changing with temperature. For the thermodynamics model, the thermal resistance-capacitance network method is used to describe the heat conduction process of each part of the cable (conductor, insulation layer, shielding layer, outer sheath, etc.). For example, for a typical XLPE-insulated power cable, it is divided into a conductor area, an insulation layer area, a shielding layer area, and an outer sheath area from the inside to the outside in sequence, and each area has specific thermal resistance and capacitance parameters. The thermal resistance parameters depend on the material thermal conductivity and geometric dimensions, and the capacitance parameters are related to the material specific heat capacity and mass. The relationship between electrical characteristics changing with temperature is mainly reflected in the temperature coefficient of the conductor resistance. The cable load-loss-temperature dynamic response equation combines the above two models to form a set of time-dependent differential equations, describing the dynamic change relationship of current, loss, and temperature with time. This system of equations considers multiple factors such as cable material characteristics, structural parameters, current changes, and heat dissipation conditions, and can accurately reflect the thermoelectric coupling behavior of the cable during actual operation. For cables in long-term service, the influence of insulation aging on the thermal conductivity also needs to be considered. Usually, the thermal conductivity of the XLPE insulation layer will decrease by about 5-8% after 10 years of aging, and this change is also integrated into the coupling relationship model.

[0035] Specifically, according to the laying method information and ambient temperature data in the topological data, the heat dissipation parameters of each section of the cable are determined and integrated into the coupling relationship model. The heat dissipation conditions of the cable directly affect its temperature rise characteristics and current-carrying capacity, and the heat dissipation conditions depend on the laying method and environmental factors. Common cable laying methods include direct burial, pipeline, cable trench, cable tunnel, overhead, etc., and each laying method corresponds to different heat dissipation characteristics. For example, the heat dissipation of directly buried cables mainly relies on the surrounding soil, and its heat dissipation parameters are related to the soil thermal resistivity (typical value is 0.8 - 2.5 K·m / W), burial depth (generally 0.7 - 1.2 m), and soil humidity; the heat dissipation of cables laid in pipelines is limited by the heat conduction of the air in the pipe, and the heat dissipation conditions are poor, usually a derating factor of 20 - 30% needs to be set; the heat dissipation of cables in cable trenches or tunnels has both convective and radiative effects, and its heat dissipation parameters are significantly affected by ventilation conditions. By analyzing the laying information recorded in the topological data, such as burial depth, pipe diameter, trench size, arrangement method, etc., and combining with the principles of thermodynamics, the environmental thermal resistance value of each section of the cable is calculated. For long-distance cables with a mixture of multiple laying methods, it is necessary to calculate and record the heat dissipation parameters of different sections separately. The ambient temperature data is obtained through a real-time monitoring system or a meteorological historical database, and the variation law of ambient temperature in different regions and seasons is recorded. For urban underground cables, the heat island effect, surrounding heat sources (such as thermal pipelines), and the thermal superposition effect of other parallel cables also need to be considered. For example, in a cable tunnel in the central area of a certain city, 8 circuits of 10 kV cables and 4 circuits of 35 kV cables are laid. During the peak load period, the temperature in the tunnel is more than 12 °C higher than the ambient temperature, and the heat dissipation conditions of each cable are severely restricted at this time. Integrating these heat dissipation parameters into the coupling relationship model requires modifying the boundary condition equation of the thermodynamic model so that the heat transfer calculation can accurately reflect the actual heat dissipation state in various laying environments, thereby improving the accuracy of the temperature distribution calculation.

[0036] Specifically, using the coupled relationship model integrated with heat dissipation parameters, the time-step rolling calculation method is adopted to iteratively solve the temperature distribution of the cable under the typical load curve. The time-step rolling calculation is a numerical method suitable for dealing with the relationship between time-varying load and temperature dynamic response. Its core idea is to discretize the continuous time process into a series of small time intervals. Within each time step, it is assumed that the load remains constant, and the temperature state at the end of this period is calculated and used as the initial condition for the next period. For power cables, 15 minutes or 30 minutes is usually selected as the basic time step to balance the calculation accuracy and calculation amount. Within each time step, it is necessary to solve the dynamic response equation of cable load-loss-temperature to obtain the changes in conductor temperature, insulation layer temperature, and outer sheath temperature. For cable systems with large thermal inertia (such as extra-high voltage cross-linked polyethylene cables), their temperature response has obvious hysteresis, and at this time, numerical integration methods such as Runge-Kutta need to be combined to improve the calculation accuracy. The selection of the typical load curve is based on the actual operation data of the power grid, such as the typical daily load curves on weekdays, holidays, and in summer and winter. For example, for the distribution cables in urban commercial areas, the typical daily load shows a "double-peak" characteristic, reaching peaks at 10-12 o'clock and 19-21 o'clock respectively. This load characteristic determines the time-varying characteristics of the cable temperature distribution. For complex cable networks with branch structures, it is necessary to consider the problem of current redistribution caused by load transfer and power flow changes, which requires collaborative iteration of power flow calculation and temperature calculation. As the calculation progresses, the cable temperature distribution gradually forms over time, reflecting the temperature change law of the cable within a day or a longer period, including key information such as temperature peaks, temperature rise rates, and temperature gradients. These temperature distribution data are directly related to the safe operation of the cable and are also the basis for subsequent accurate loss calculation.

[0037] Specifically, based on the calculated temperature distribution and the resistance-temperature coefficient relationship in the coupling relationship model, the actual resistance values of the cables in each period are updated, and the time-varying loss data is calculated and output. The relationship between the cable conductor resistance and temperature is accurately described by the resistance-temperature coefficient. As the temperature increases, the actual resistance value of the cable increases significantly. Using the linear temperature coefficient model mentioned above, for each cable segment in the cable network, the corresponding actual resistance value is calculated according to the conductor temperature in this period. For cables in long-term operation, the influence of the increased contact resistance at the joints also needs to be considered. Usually, for cables with joints aged over 20 years, the resistance increment at the joints can reach 5-10%. The updated resistance value is combined with the current value to recalculate the cable loss: P = I²·R(T), where R(T) is the resistance value considering the influence of the actual temperature T, which is much more accurate than the rated resistance used in the initial calculation. In addition to the conductor loss, the complete cable loss also includes dielectric loss and shield loss. The dielectric loss is proportional to the square of the voltage and the tangent value of the dielectric loss angle. The tangent value of the dielectric loss angle also changes with temperature. Usually, for every 10°C increase in temperature, the dielectric loss will increase by 15-20%. The shield loss is mainly generated by the induced current and is related to the main current, the shield impedance, and the electromagnetic coupling coefficient. Combining these three parts of losses, the total loss value of each cable segment in different periods is formed, constituting a complete time-varying loss data matrix. This matrix contains the loss distribution information in both the time dimension and the space dimension, comprehensively reflecting the dynamic change characteristics of the losses in the cable network. By analyzing the time-varying loss data, the loss hotspots and high-loss periods in the cable network can be identified. For example, the time-varying loss analysis of a 35 kV cable network in a certain area shows that the loss value of an aging cable in the load concentration area is 35% higher than the normal value in the afternoon of summer, becoming an obvious hotspot in the network. Such information has important guiding value for the power grid planning and operation, providing data support for the subsequent optimization decision-making. The time-varying loss data not only reflects the working state of the cable but also indirectly indicates the aging degree and potential risks of the cable Further, the step of updating the actual resistance values of the cables in each period according to the calculated temperature distribution and the resistance-temperature coefficient relationship in the coupling relationship model, and calculating and outputting the time-varying loss data includes: calculating the actual resistance value of each cable segment in each period in the cable network according to the calculated temperature distribution; substituting the actual resistance value into Ohm's law and combining with the current value in the voltage power distribution data to calculate the line loss data of the cables in each period; identifying the cable joint positions according to the laying method information, calculating the additional losses at the joint positions, and incorporating the additional losses into the line loss data; performing time series integration on the line loss data in each period to obtain a time-varying loss data matrix, and calculating the average loss level, the loss change characteristics over time, and the loss spatial distribution characteristics of each cable segment based on the time-varying loss data matrix, and outputting the time-varying loss data of the cable network

[0038] Specifically, according to the calculated temperature distribution and the resistance temperature coefficient relationship in the coupling relationship model, the process of updating the actual resistance value of the cable in each time period and calculating and outputting the time-varying loss data first requires calculating the actual resistance value of each section of the cable in the cable network for each time period according to the calculated temperature distribution. In specific operations, using the conductor temperature distribution data calculated in the foregoing steps, for the temperature value at each time point, the resistance value is updated by applying the resistance temperature coefficient relationship. The resistance temperature coefficient relationship is a basic physical model that describes the law of change of the resistance of a metal conductor with temperature. For the commonly used copper and aluminum conductors of power cables, their resistance values are approximately linearly related to the temperature. When calculating the resistance, the formula R(T) = R 20 [1 + α(T - 20)] is used, where R 20 is the conductor resistance at the standard temperature of 20°C, α is the temperature coefficient of the conductor material, approximately 0.00393 / °C for copper conductors and approximately 0.00403 / °C for aluminum conductors, and T is the actual conductor temperature in the current calculation period. For example, for a section of YJV22-3×150 cable in a 10 kV distribution network with a length of 0.7 km, the resistance per phase at 20°C is 0.124 Ω / km. If the conductor temperature calculated in a certain time period is 65°C, then the actual resistance value at this time is updated to 0.124×[1 + 0.00393×(65 - 20)]×0.7 = 0.0996 Ω, which is approximately 17.6% higher than the resistance at the standard temperature. For cables with a longer service life, the oxidation of the conductor surface will increase the contact resistance. Therefore, the resistance temperature coefficient also needs to be corrected according to the service life of the cable. Usually, for copper conductor cables that have been used for more than 15 years, their resistance temperature coefficient α will increase by 2-5%. Through this method, the actual resistance values of each section of the cable in the cable network at each time point are updated to form a complete time-varying resistance matrix, which accurately reflects the electrical characteristics of the cable at the actual operating temperature.

[0039] Specifically, substitute the actual resistance value into Ohm's law and calculate the line loss data of the cable for each time period in combination with the current value in the voltage-power distribution data. In this step, using the updated time-varying resistance matrix and the current distribution obtained from the power flow calculation, calculate the conductor loss of each section of the cable network at each time point. The conductor loss calculation uses the formula P = I²·R, where I is the effective current value and R is the actual resistance value at the current temperature. Since the cable current varies with the load and the resistance varies with the temperature, the conductor loss has obvious time-varying characteristics. For example, for a section of ZRC-YJLW03-1×400 cable in a 35 kV distribution network in a commercial area of a certain city, the current reaches 320 A and the temperature rises to 72 °C during the peak load period on weekdays (about 14:00-16:00). At this time, the calculated actual resistance is 0.0872 Ω / km, and the conductor loss is about 8.92 kW / km; while during the light load period at night (about 02:00-04:00), the current drops to 105 A, the temperature drops to 35 °C, the actual resistance is 0.0759 Ω / km, and the conductor loss is only 0.84 kW / km, with a difference of more than 10 times. For three-phase cables, additional losses caused by unbalanced three-phase currents need to be considered. At this time, the current and resistance of each phase need to be calculated separately and then summed. For long-distance large-section cables, the skin effect and proximity effect will cause an increase in high-frequency losses, especially in load areas with more harmonic components, and corresponding corrections need to be added on the basis of the basic loss calculation. By calculating the losses of each section of the cable at each time point, complete line loss data of the cable network is obtained. This data contains distribution information in both the time and space dimensions and can accurately reflect the distribution characteristics and change rules of losses in the power grid.

[0040] Specifically, identify the cable joint positions based on the laying method information, calculate the additional losses at the joint positions, and incorporate the additional losses into the line loss data. Cable joints are special parts in the cable line. Due to their complex structure and high technological requirements, there are often additional contact resistances and dielectric losses at the joints, making them the weak links and loss hotspots in the cable line. The identification of cable joint positions mainly relies on the connection records and length information in the cable laying data. Usually, cable joints appear in the following positions: at the connections of the standard cable lengths (such as 500 meters or 1000 meters), at the turning points where the cable crosses buildings or roads, at the connections of cables laid in different models or at different times, and at the connections of the cable repair and replacement sections. For the calculation of the additional losses at the joints, consider the increase in contact resistance at the joints and the loss changes caused by local structural changes. Generally speaking, for newly installed crimp-type cable joints, the increase in contact resistance is about 0.1 - 0.2 times the resistance per unit length of the cable; for joints that have been used for more than 5 years, the increase in contact resistance may reach 0.3 - 0.5 times; for old joints that have been used for more than 15 years, especially those that have experienced overload or short-circuit impacts, the increase in contact resistance may be as high as 0.8 - 1.2 times, becoming significant loss hotspots. For example, in an 11 kV distribution line, there is an intermediate joint that has been in operation for 18 years. Through infrared temperature measurement, it is found that the temperature at this point is 15 °C higher than that of the surrounding cables. Through calculation, it is determined that its additional losses are about 1.7 times that of the normal section. Incorporating the additional losses at the identified joint positions into the overall line loss data to form a more complete and accurate cable network loss distribution map is of great significance for identifying risk points in the power grid and optimizing cable operation.

[0041] Specifically, time - series integration is performed on the line loss data of each time period to obtain a time - varying loss data matrix. Based on the time - varying loss data matrix, the average loss level, the characteristics of loss variation over time, and the characteristics of loss spatial distribution of each cable section are calculated, and the time - varying loss data of the cable network is output. Time - series integration organizes the line loss data at discrete time points into a continuous time - series data structure for subsequent analysis and application. During the integration process, the loss data of each cable section in the cable network at each calculation time period is arranged in chronological order to form a time - varying loss data matrix. The rows of this matrix represent cable sections, the columns represent time points, and the matrix element values are the loss values at the corresponding space - time points. Based on this matrix, various statistical characteristic quantities are calculated. For example, the average loss level of each cable section (corresponding to the average value of the matrix row) reflects the basic loss condition of the cable; the characteristics of loss variation over time (corresponding to the change trend of the matrix column) reflect the correlation between loss and load fluctuation; the characteristics of loss spatial distribution (corresponding to the comparison between matrix rows) reveal the non - uniformity of loss distribution and the hot - spot positions in the network. For example, by analyzing the time - varying loss data matrix of a 110 kV cable loop network in the core area of a certain city, it is found that a 1.2 - kilometer - long cable located at the load center has a daily average loss as high as 6.8 kW / km, which is 2.3 times the network average level, and the loss peak reaches 11.5 kW / km during the high - temperature period in summer, showing obvious spatio - temporal aggregation characteristics. Through further analysis, it is determined that this abnormal phenomenon is comprehensively related to the aging degree, laying environment, and load characteristics of this section of the cable. In addition, the time - varying loss data matrix can also generate various visualization charts, such as loss heat maps, time - varying curve charts, and spatial distribution charts, etc., to intuitively display the distribution law of the cable network loss.

[0042] 104. Perform multi - dimensional feature analysis on the time - varying loss data, calculate the cable loss risk index through multi - dimensional feature analysis, identify loss hot - spots according to the loss risk index, and generate a cable optimization plan.

[0043] In an embodiment of the present invention, the performing multi - dimensional feature analysis on the time - varying loss data, calculating the cable loss risk index through the multi - dimensional feature analysis, identifying loss hot - spots according to the loss risk index, and generating a cable optimization plan includes: extracting characteristic parameters of three dimensions, namely, the time dimension, the space dimension, and the working condition dimension, from the time - varying loss data; calculating the cable loss risk index of each cable section through a feature fusion algorithm according to the characteristic parameters of the three dimensions; based on the cable loss risk index and a preset risk threshold, identifying the cable sections with loss risks exceeding the threshold as loss hot - spots, analyzing the distribution of the characteristic parameters of the loss hot - spots, determining the dominant characteristic parameter combination of each loss hot - spot, and defining the dominant characteristic parameter combination as the characteristic pattern of the loss hot - spot; for the loss hot - spots with different characteristic patterns, generating corresponding cable optimization plans through a multi - level cable optimization decision tree based on the topological data.

[0044] Specifically, a multi-dimensional feature analysis is performed on the time-varying loss data, and the cable loss risk index is calculated through the multi-dimensional feature analysis. The implementation process of identifying the loss hotspots and generating the cable optimization plan according to the loss risk index first involves extracting the characteristic parameters of the three dimensions of time dimension, space dimension and working condition dimension from the time-varying loss data. In the time dimension, the extracted characteristic parameters mainly include the daily loss peak-to-valley ratio, the loss duration ratio and the loss fluctuation coefficient. The daily loss peak-to-valley ratio reflects the ratio of the maximum value to the minimum value of the cable loss in a day. Generally, the larger the peak-to-valley ratio, the more violent the cable loss fluctuation with the load, and the more obvious the thermal stress shock. The loss duration ratio refers to the proportion of the cumulative time when the cable loss exceeds the preset threshold (such as 80% of the rated loss) to the whole day. The larger the value, the longer the high loss state of the cable lasts, and the more significant the thermal accumulation effect. The loss fluctuation coefficient is the ratio of the loss standard deviation to the average value, which reflects the stability of the loss change. In the spatial dimension, the extracted characteristic parameters include the relative position coefficient, the local loss gradient and the neighboring influence factor. The relative position coefficient indicates the position characteristics of the cable segment in the entire network topology, such as being located at the load center, near the power source, or at the end; the local loss gradient describes the rate of change of the cable loss in space. The larger the gradient, the more uneven the local loss distribution; the neighboring influence factor quantifies the degree of influence of the surrounding cables or heat sources on the heat dissipation conditions of the target cable. In the working condition dimension, the extracted characteristic parameters include load sensitivity coefficient, ambient temperature influence coefficient, and seasonal variation coefficient. The load sensitivity coefficient indicates the degree of response of the loss to load changes; the ambient temperature influence coefficient describes the intensity of the influence of ambient temperature changes on cable losses; and the seasonal variation coefficient reflects the fluctuation of losses in different seasons. Through the feature extraction of these three dimensions, the time-varying characteristics, spatial distribution characteristics, and working condition dependence of cable losses are fully characterized, providing a multi-angle data basis for subsequent risk assessment.

[0045] Specifically, based on the extracted characteristic parameters in three dimensions, the cable loss risk index of each cable section is calculated through a feature fusion algorithm. Feature fusion is the process of integrating characteristic parameters with different dimensions and different dimensions into a single evaluation index. In practical applications, a hierarchical weighted fusion method is adopted. First, feature normalization is performed within each dimension to convert parameters with different dimensions into unified 0-1 interval values. Then, weights are assigned according to the degree of influence of each parameter on the loss risk, and the sub-index within each dimension is calculated. For the time dimension, the calculation of its sub-index focuses on the time distribution characteristics of the loss, such as the loss level, duration, and fluctuation amplitude during peak periods; for the spatial dimension, the calculation of its sub-index focuses on the location characteristics of the cable, the risk of local heat accumulation, and the influence of the surrounding environment; for the operating condition dimension, its sub-index focuses on evaluating the loss response sensitivity of the cable under different load and environmental conditions. After the calculation of the sub-indexes in the three dimensions is completed, the weight coefficients of each dimension in the total risk index are determined through an adaptive weight method. The adaptive weight method dynamically adjusts the importance of each dimension according to the operating characteristics of the power grid and historical fault data. For example, in the core urban area with intensive load, the weight of the spatial dimension is often higher; for the industrial area with large load fluctuations, the weight of the time dimension may be greater; for the tourist area with obvious seasonal load, the weight of the operating condition dimension is relatively prominent. Through the weighted combination of the sub-indexes in the three dimensions, the comprehensive loss risk index of each cable section is finally calculated. The value of this index ranges from 0 to 10, and the larger the value, the higher the loss risk. For example, for a severely aged cable section in a certain urban distribution network, its sub-index in the time dimension is 8.2, its sub-index in the spatial dimension is 7.5, and its sub-index in the operating condition dimension is 6.8. The comprehensive risk index after weighted fusion is 7.6, which is at a high risk level.

[0046] Specifically, based on the calculated cable loss risk index and the preset risk threshold, identify the cable segments with loss risks exceeding the threshold as loss hotspots, analyze the distribution of the characteristic parameters of the loss hotspots, determine the dominant characteristic parameter combination for each loss hotspot, and define the dominant characteristic parameter combination as the characteristic pattern of the loss hotspot. The setting of the risk threshold is usually based on the grid safe operation standards and historical experience data. Generally, the risk index is divided into four levels: low risk (0 - 3), medium risk (3 - 6), high risk (6 - 8), and extremely high risk (8 - 10), and the cable segments with high risk and extremely high risk are identified as loss hotspots. For each identified loss hotspot, further analyze the distribution of its characteristic parameters in each dimension to determine the main cause of the risk. This analysis uses the characteristic contribution calculation method to calculate the contribution rate of each characteristic parameter to the risk index. The characteristic parameters with a contribution rate exceeding the preset threshold (such as 20%) are regarded as the dominant characteristic parameters. According to the combination method of the dominant characteristic parameters, the loss hotspots are classified into different characteristic patterns. Typical characteristic patterns include: time - dominant type (such as excessive peak - valley ratio, long high - loss duration, etc.), space - dominant type (such as located in the heat - source - intensive area, poor local heat dissipation conditions, etc.), operating condition - dominant type (such as large load fluctuations, high seasonal sensitivity, etc.), and mixed type (the combined action of multi - dimensional factors). For example, for a loss hotspot with a risk index of 8.3, it is found that the contribution rates of its loss duration ratio and loss fluctuation coefficient are 25% and 22% respectively, while the contribution rate of the relative position coefficient is 30%, and the contribution rates of other parameters are all lower than 15%. Therefore, this hotspot is defined as the "time - space mixed type" characteristic pattern, and its main feature is the risk of heat accumulation caused by the superposition of a long - time high - loss state and an unfavorable spatial position. Through this characteristic pattern division, the internal mechanism of the formation of loss hotspots can be deeply understood, providing targeted guidance for the formulation of subsequent optimization measures.

[0047] Specifically, for the loss hotspots with different characteristic patterns, according to the topological data, a corresponding cable optimization scheme is generated through a multi-level cable optimization decision tree. The multi-level cable optimization decision tree is a structured decision support model. According to the characteristic patterns of the loss hotspots and the characteristics of the network topology, the most suitable optimization scheme is screened layer by layer according to the preset decision rules. The root node of the decision tree is the characteristic pattern of the loss hotspot, the first-layer branches are the selection of optimization directions, the second-layer branches are the selection of specific measures, and the leaf nodes are the final optimization schemes. For time-dominated hotspots, the optimization directions mainly focus on load adjustment and current balance, and the specific measures include load transfer, adjustment of time-of-use power supply strategies, three-phase load balance, etc.; for space-dominated hotspots, the optimization directions focus on improving the heat dissipation conditions and cable replacement, and the specific measures include adjusting the laying method, adding heat dissipation devices, replacing large-section cables or segmented composite cables with different cross-sections, etc.; for condition-dominated hotspots, the optimization directions focus on enhancing adaptability and monitoring capabilities, and the specific measures include installing intelligent monitoring devices, implementing dynamic capacity management, adjusting seasonal operation modes, etc.; for mixed hotspots, multiple measures need to be comprehensively considered and prioritized according to the weight of the dominant factors. For example, for the aforementioned "time-space mixed" hotspot, the decision tree first selects the optimization direction of "load adjustment + heat dissipation improvement", and then recommends a combination of measures of "time-of-use load transfer + local modification of the laying method" according to the cable location and load characteristics, and finally forms a detailed optimization scheme, including the specific load transfer ratio, implementation time period, and laying modification technical requirements, etc. For each optimization scheme, the implementation cost and expected loss reduction benefits are also calculated, and the investment payback period is calculated to provide an economic evaluation basis for the final decision.

[0048] In this embodiment, by obtaining the cable topology data, extracting the parameter characteristics, and generating a parameter database; based on the parameter database, a matrix transformation is performed to establish a network calculation model, a power flow calculation is executed, and voltage-power distribution data is output; according to the voltage-power distribution data, a coupling relationship model between temperature and loss is established, the cable temperature distribution is calculated, and time-varying loss data is generated; a multi-dimensional feature analysis is performed on the time-varying loss data, the loss risk index is calculated, the loss hotspots are identified, and an optimization scheme is generated. The present invention realizes the accurate analysis of the power grid cable loss by establishing a dynamic relationship model between the cable parameters and the loss, combining the characteristics of the cable shielding layer and the temperature change factors, provides technical support for identifying the loss hotspots and formulating targeted optimization schemes, and effectively reduces the calculation deviation of the power grid loss.

[0049] The method for analyzing the power grid cable loss in the embodiment of the present invention has been described above. Next, the power grid cable loss analysis device in the embodiment of the present invention will be described. The power grid cable loss analysis device is shown in Figure 2 , and an embodiment of the power grid cable loss analysis device in the embodiment of the present invention includes: The data acquisition module 201 is used to obtain the topological data of the cables in the power grid system, extract the cable parameter characteristics according to the topological data, and generate a cable parameter database; The power flow calculation module 202 is used to perform parameter matrix conversion on the cable network in the power grid system based on the cable parameter database and establish a network calculation model, execute power flow calculation through the network calculation model, and output the voltage-power distribution data of the cable network; The loss calculation module 203 is used to establish a coupling relationship model between the cable temperature and the loss according to the voltage-power distribution data, calculate the temperature distribution of the cable under different conditions through the coupling relationship model, and generate the time-varying loss data of the cable network; The hot spot identification module 204 is used to perform multi-dimensional feature analysis on the time-varying loss data, calculate the cable loss risk index through the multi-dimensional feature analysis, identify the loss hot spots according to the loss risk index, and generate a cable optimization plan.

[0050] In the embodiment of the present invention, the power grid cable loss analysis device runs the above-mentioned power grid cable loss analysis method. The power grid cable loss analysis device obtains cable topological data, extracts parameter characteristics, and generates a parameter database; performs matrix conversion based on the parameter database to establish a network calculation model, executes power flow calculation, and outputs voltage-power distribution data; establishes a coupling relationship model between temperature and loss according to the voltage-power distribution data, calculates the cable temperature distribution, and generates time-varying loss data; performs multi-dimensional feature analysis on the time-varying loss data, calculates the loss risk index, identifies the loss hot spots, and generates an optimization plan. The present invention realizes the accurate analysis of the power grid cable loss by establishing a dynamic relationship model between the cable parameters and the loss, combining the characteristics of the cable shielding layer and the temperature change factors, provides technical support for identifying the loss hot spots and formulating targeted optimization plans, and effectively reduces the calculation deviation of the power grid loss.

[0051] The present invention also provides a computer-readable storage medium. The computer-readable storage medium can be a non-volatile computer-readable storage medium, or can also be a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions run on a computer, the computer is made to execute the steps of the power grid cable loss analysis method.

[0052] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the above-described system or device and unit can refer to the corresponding processes in the foregoing method embodiments, and will not be described herein again.

[0053] When the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0054] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A method for analyzing power grid cable losses, characterized in that, The power grid cable loss analysis method includes: Obtaining the topological data of the cables in the power grid system, extracting the cable parameter characteristics according to the topological data, and generating a cable parameter database; Based on the cable parameter database, performing parameter matrix conversion on the cable network in the power grid system and establishing a network calculation model, executing power flow calculation through the network calculation model, and outputting the voltage-power distribution data of the cable network; According to the voltage-power distribution data, establishing a coupling relationship model between cable temperature and loss, calculating the temperature distribution of the cable under different conditions through the coupling relationship model, and generating the time-varying loss data of the cable network; Performing multi-dimensional feature analysis on the time-varying loss data, calculating the cable loss risk index through the multi-dimensional feature analysis, identifying the loss hotspots according to the loss risk index, and generating a cable optimization plan.

2. The power grid cable loss analysis method according to claim 1, wherein The obtaining the topological data of the cables in the power grid system, extracting the cable parameter characteristics according to the topological data, and generating a cable parameter database includes: Collecting the connection relationship data between the cables through the automation equipment of the power system to form topological data; Analyzing the grounding method of the cable shielding layer according to the topological data, and calculating the corresponding shielding layer current path coefficient for various grounding methods; Based on the cable connection relationship data and the service life records of the cables, assigning aging degradation parameters to each cable; Using the shielding layer current path coefficient and the aging degradation parameters, calculating the equivalent impedance values corresponding to each parameter combination through electromagnetic field numerical simulation, and generating a cable parameter database.

3. The power grid cable loss analysis method according to claim 1, wherein The based on the cable parameter database, performing parameter matrix conversion on the cable network in the power grid system and establishing a network calculation model, executing power flow calculation through the network calculation model, and outputting the voltage-power distribution data of the cable network includes: Constructing a node-branch incidence matrix of the cable network according to the cable parameter database, and dividing the cable network into two types of ports, namely load nodes and power supply nodes, based on the node-branch incidence matrix; Constructing a hybrid parameter matrix based on the two types of ports and the equivalent impedance values in the cable parameter database, and converting the hybrid parameter matrix into an admittance matrix through matrix operation; Calculating the correction factor of the shielding layer circulating current according to the cable shielding layer information in the cable parameter database, applying the correction factor to the admittance matrix, and generating a corrected admittance matrix, and the corrected admittance matrix constitutes the network calculation model; Using the corrected admittance matrix in the network calculation model to execute power flow calculation by applying the improved Newton-Raphson iteration algorithm to obtain the voltage-power distribution data of the cable network.

4. The power grid cable loss analysis method according to claim 3, wherein The using the corrected admittance matrix in the network calculation model to execute power flow calculation by applying the improved Newton-Raphson iteration algorithm to obtain the voltage-power distribution data of the cable network includes: Constructing a node voltage balance equation set for power flow calculation according to the corrected admittance matrix, and setting an initial voltage value for the node voltage balance equation set as the iteration starting point; Calculating the Jacobian matrix for the node voltage balance equation set, and solving the power flow node voltage correction amount by using the improved Newton-Raphson iteration method based on the Jacobian matrix; Update the node voltage value in the power flow calculation according to the node voltage correction amount, and judge the power flow convergence of the node voltage value. If the power flow convergence condition is not reached, return to the previous step and continue the iterative calculation; Using the node voltage value after power flow convergence, calculate the power flow current distribution of each branch in the cable network according to Ohm's law; According to the node voltage value and the power flow current distribution, calculate the power flow power distribution of each section of the cable in the cable network to obtain the voltage-power distribution data of the cable network.

5. The power grid cable loss analysis method according to claim 1, characterized in that Based on the voltage-power distribution data, establish a coupling relationship model between cable temperature and loss, and calculate the temperature distribution of the cable under different conditions through the coupling relationship model to generate the time-varying loss data of the cable network, including: Calculate the current value and initial loss value of each section of the cable according to the voltage-power distribution data; Based on the current value and the initial loss value, construct a coupling relationship model between cable temperature and loss that includes the mutual influence of current heating, temperature rise, and resistance change. The coupling relationship model is characterized by a cable load-loss-temperature dynamic response equation; According to the laying method information and ambient temperature data in the topology data, determine the heat dissipation parameters of each section of the cable, and integrate the heat dissipation parameters into the coupling relationship model; Using the coupling relationship model integrated with heat dissipation parameters, adopt a time-step rolling calculation method to iteratively solve the temperature distribution of the cable under a typical load curve; Based on the calculated temperature distribution and the resistance-temperature coefficient relationship in the coupling relationship model, update the actual resistance value of the cable in each time period, and calculate and output the time-varying loss data.

6. The power grid cable loss analysis method according to claim 5, characterized in that, The updating of the actual resistance value of the cable in each time period based on the calculated temperature distribution and the resistance-temperature coefficient relationship in the coupling relationship model, and the calculation and output of the time-varying loss data include: According to the calculated temperature distribution, calculate the actual resistance value of each section of the cable in the cable network for each time period; Substitute the actual resistance value into Ohm's law, and combine with the current value in the voltage-power distribution data to calculate the line loss data of the cable in each time period; Identify the cable joint positions according to the laying method information, calculate the additional loss at the joint positions, and incorporate the additional loss into the line loss data; Perform time series integration on the line loss data in each time period to obtain a time-varying loss data matrix, and calculate the average loss level, loss time-varying characteristics, and loss spatial distribution characteristics of each cable section based on the time-varying loss data matrix, and output the time-varying loss data of the cable network.

7. The method for analyzing the loss of power grid cables according to claim 1, characterized in that, The multi-dimensional feature analysis of the time-varying loss data, calculating the cable loss risk index through the multi-dimensional feature analysis, and identifying the loss hotspots and generating the cable optimization plan according to the loss risk index include: Extract the characteristic parameters of the three dimensions of time dimension, space dimension, and working condition dimension from the time-varying loss data respectively; According to the characteristic parameters of the three dimensions, calculate the cable loss risk index of each section of the cable through a feature fusion algorithm; Based on the cable loss risk index and a preset risk threshold, identify the cable segments with loss risks exceeding the threshold as loss hotspots, analyze the distribution of characteristic parameters of the loss hotspots, determine the dominant characteristic parameter combination for each loss hotspot, and define the dominant characteristic parameter combination as the characteristic pattern of the loss hotspot; For the loss hotspots with different characteristic patterns, according to the topological data, generate corresponding cable optimization schemes through a multi-level cable optimization decision tree.

8. A power grid cable loss analysis device, characterized in that, The power grid cable loss analysis device includes: A data acquisition module, configured to obtain the topological data of the cables in the power grid system, extract the cable parameter characteristics according to the topological data, and generate a cable parameter database; A power flow calculation module, configured to perform parameter matrix conversion on the cable network in the power grid system based on the cable parameter database and establish a network calculation model, execute power flow calculation through the network calculation model, and output the voltage power distribution data of the cable network; A loss calculation module, configured to establish a coupling relationship model between cable temperature and loss according to the voltage power distribution data, calculate the temperature distribution of the cable under different conditions through the coupling relationship model, and generate the time-varying loss data of the cable network; A hotspot identification module, configured to perform multi-dimensional feature analysis on the time-varying loss data, calculate the cable loss risk index through the multi-dimensional feature analysis, identify loss hotspots according to the loss risk index, and generate a cable optimization scheme.

9. A computer-readable storage medium having instructions stored thereon, characterized in that, When the instruction is executed by the processor, the steps of the power grid cable loss analysis method according to any one of claims 1-7 are implemented.

Images