Power quality monitoring and intelligent optimization method and system driven by digital twin

Abstract

The present invention provides a digital twin-driven power quality monitoring and intelligent optimization method and system, which relates to the field of electric power technology, including constructing a three-dimensional geometric model, a material property model and a physical property model of a transmission line, evaluating the health status of the power quality of the transmission line through an intelligent algorithm, generating a power quality assessment report and uploading it to a cloud server; the cloud server receives the power quality assessment report uploaded by an edge computing server, and if the power quality assessment report shows that the power quality does not meet the standard, the digital twin model is called, and the optimal transmission line operating parameter combination that meets the power quality standard is searched by a multi-objective optimization algorithm through the adjustment of the digital twin model parameters; the optimal transmission line operating parameter combination optimized by the digital twin model is sent to the intelligent control system of the transmission line.

Description

Technical Field

[0001] The present invention relates to electric power technology, and in particular to a method and system for monitoring and intelligently optimizing power quality driven by digital twins. Background Art

[0002] Traditional power quality monitoring and management methods have the following shortcomings:

[0003] The monitoring means are backward. At present, power quality monitoring mainly relies on a small number of power quality analysis instruments installed at key measuring points, which makes it difficult to fully perceive the power quality status of the entire network. The monitoring data of each measuring point is stored independently, lacking fusion analysis, and it is difficult to reveal the temporal and spatial propagation laws and impact range of disturbance events. Lack of real-time decision-making ability. In the face of complex and changeable power quality disturbances, it is difficult to meet the needs of real-time control by relying on manual analysis and qualitative judgment. At present, there is a lack of intelligent optimization algorithms driven by power quality data, and it is impossible to automatically generate the best governance measures. Improper configuration of governance devices. Power quality governance equipment such as dynamic reactive power compensation devices and active power filters lack quantitative indicators in capacity planning and installation location selection, making it difficult to adapt to rapidly changing grid operating conditions, resulting in investment waste and poor governance effects. Lack of global collaborative optimization. Various power quality governance equipment is independently controlled, lacks unified coordination, and it is difficult to achieve optimal operation within the entire system. There is a lack of coordination between control measures at different time scales, such as the separation of links such as rapid fault location and isolation, voltage reactive power optimization, and harmonic suppression. Summary of the invention

[0004] The embodiments of the present invention provide a digital twin-driven power quality monitoring and intelligent optimization method and system, which can solve the problems in the prior art.

[0005] According to a first aspect of the embodiments of the present invention,

[0006] Provides digital twin-driven power quality monitoring and intelligent optimization methods, including:

[0007] Construct a three-dimensional geometric model, material property model and physical property model of the transmission line, and establish a digital twin model of the transmission line by fusing multi-source heterogeneous data; wherein the three-dimensional geometric model includes geometric parameters of the tower, conductor, insulator string and hardware transmission line, the material property model includes the material model of the conductor, the material parameters of the dielectric constant and dissipation factor of the insulating material, and the physical property model includes the electrical parameters of the resistivity, inductance and capacitance of the conductor, and the physical parameters of the mechanical strength of the tower and insulator string; arrange a power quality monitoring unit at the measuring point of the transmission line, the power quality monitoring unit includes a high-precision voltage transformer, a current transformer, a harmonic analyzer, and a temperature and humidity sensor, and the power quality monitoring unit uploads the collected and analyzed power quality data to the edge computing server in real time through the optical fiber private network; the edge computing server pre-processes the power quality data, evaluates the health status of the power quality of the transmission line through an intelligent algorithm, generates a power quality evaluation report and uploads it to the cloud server;

[0008] The cloud server receives the power quality assessment report uploaded by the edge computing server. If the power quality assessment report shows that the power quality does not meet the standard, the digital twin model is called to simulate the power quality optimization process of the transmission line in the virtual space by adjusting the parameters of the digital twin model. On the basis of solving the power flow equation group, a multi-objective optimization algorithm is used to search for the optimal transmission line operation parameter combination that meets the power quality standard, including the conductor cross-sectional area, reactive power compensation capacity, and ambient temperature and humidity control threshold.

[0009] The optimal transmission line operating parameter combination obtained by optimizing the digital twin model is sent to the intelligent control system of the transmission line. The intelligent control system collaboratively controls the reactive compensation device, the dynamic voltage regulator, and the intelligent de-icing device, and dynamically adjusts the reactive power, voltage amplitude and phase, and conductor temperature parameters of the transmission line, so that the transmission line reaches the optimal operating state under the current environmental conditions and meets the power quality requirements. The intelligent control system feeds back the optimized control effect to the digital twin model for correcting and updating the relevant parameters of the digital twin model.

[0010] According to a second aspect of the embodiments of the present invention,

[0011] Provide digital twin-driven power quality monitoring and intelligent optimization system, including:

[0012] The first unit is used to construct a three-dimensional geometric model, material attribute model and physical property model of the transmission line, and establish a digital twin model of the transmission line by fusing multi-source heterogeneous data; wherein the three-dimensional geometric model includes the geometric parameters of the tower, conductor, insulator string and hardware transmission line, the material attribute model includes the material model of the conductor, the material parameters of the dielectric constant and dissipation factor of the insulating material, and the physical property model includes the electrical parameters of the resistivity, inductance and capacitance of the conductor, and the physical parameters of the mechanical strength of the tower and insulator string; a power quality monitoring unit is arranged at the measuring point of the transmission line, and the power quality monitoring unit includes a high-precision voltage transformer, a current transformer, a harmonic analyzer, and a temperature and humidity sensor. The power quality monitoring unit uploads the power quality data collected and analyzed to the edge computing server in real time through the optical fiber private network; the edge computing server pre-processes the power quality data, evaluates the health status of the power quality of the transmission line through an intelligent algorithm, generates a power quality evaluation report and uploads it to the cloud server;

[0013] The second unit is used for the cloud server to receive the power quality assessment report uploaded by the edge computing server. If the power quality assessment report shows that the power quality does not meet the standard, the digital twin model is called to simulate the process of power quality optimization of the transmission line in the virtual space by adjusting the parameters of the digital twin model. On the basis of solving the power flow equation group, a multi-objective optimization algorithm is used to search for the optimal combination of transmission line operation parameters that meet the power quality standard, including the conductor cross-sectional area, reactive power compensation capacity, and ambient temperature and humidity control threshold;

[0014] The third unit is used to send the optimal transmission line operating parameter combination obtained by optimizing the digital twin model to the intelligent control system of the transmission line. The intelligent control system collaboratively controls the reactive compensation device, the dynamic voltage regulator, and the intelligent de-icing device, and dynamically adjusts the reactive power, voltage amplitude and phase, and conductor temperature parameters of the transmission line to enable the transmission line to achieve the optimal operating state under the current environmental conditions and meet the power quality requirements. The intelligent control system feeds back the optimized control effect to the digital twin model for correcting and updating the relevant parameters of the digital twin model.

[0015] According to a third aspect of the embodiments of the present invention,

[0016] An electronic device is provided, comprising:

[0017] processor;

[0018] a memory for storing processor-executable instructions;

[0019] The processor is configured to call the instructions stored in the memory to execute the aforementioned method.

[0020] This paper proposes an intelligent optimization control method for transmission lines based on digital twins. This method receives the optimal control parameters output by the digital twin model and carries out multi-level collaborative control from the master station layer, substation layer, and equipment layer. The master station layer is responsible for parsing and issuing control instructions, the substation layer formulates a multi-device collaborative control plan, and the equipment layer is responsible for physical execution optimization control. Intelligent sensors sense the line status and equipment conditions in real time, and complete data aggregation and upload through the communication network. The power quality management system continuously evaluates the power quality level of the line, triggers the re-optimization of the digital twin model when an abnormality occurs, and realizes the continuous optimal control of the transmission line driven by the digital twin through equipment status mapping and closed-loop feedback of optimization results, ensuring that the power quality of the line meets the standard and the comprehensive benefits are maximized. The next research focus is to further simplify the optimization control model, improve the real-time and interpretability of the control instructions, explore the panoramic collaborative optimization control method of the line combined with digital twins and intelligent operation and inspection, and build an integrated simulation verification platform for digital twin models and line physical equipment to promote the application of the method in actual engineering.

[0021] This paper proposes a distributed collaborative optimization control strategy for transmission lines based on multi-agent game. According to the geographical and electrical characteristics of the line, multiple control zones are divided to construct a three-layer agent hierarchical architecture of "equipment-region-line", in which the regional control agent is responsible for parsing the line-level dispatching instructions to form a partitioned multi-equipment optimization strategy, which is executed in a closed loop through the equipment control agent. The line control agent evaluates the global power quality level through cross-regional information interaction, and forms collaborative control instructions to guide regional control behavior. The game theory mechanism is introduced to model the distributed collaborative optimization game process of multi-regional control agents. By designing a comprehensive utility function of "self-performance and global contribution", the game parties are encouraged to form a mutually beneficial and win-win equilibrium strategy in autonomous evolution. Based on blockchain, a trusted environment for agent strategy interaction is constructed to ensure the integrity and credibility of multi-regional collaborative games, and realize the distributed collaborative optimization control of transmission lines with emergent characteristics, which is self-organizing, adaptive, and self-evolving. Subsequent research directions mainly include: further optimizing cross-regional coordination mechanisms and shortening the strategy convergence time of multi-agents; exploring the application of deep reinforcement learning in utility function design and improving the stability of equilibrium strategies; combining sociological mechanisms to design reasonable incentives and build a credible cross-regional benefit distribution and compensation mechanism. At the same time, it is also necessary to carry out hardware and software integration testing to evaluate the engineering applicability of the proposed distributed collaborative optimization control method and provide theoretical and technical support for the large-scale deployment of intelligent transmission lines. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 A schematic diagram of a flow chart of a power quality monitoring and intelligent optimization method driven by a digital twin according to an embodiment of the present invention;

[0023] Figure 2This is a structural schematic diagram of the power quality monitoring and intelligent optimization system driven by the digital twin according to an embodiment of the present invention. DETAILED DESCRIPTION

[0024] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0025] The technical solution of the present invention is described in detail with specific embodiments below. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0026] Figure 1 Schematic diagram of the process of the power quality monitoring and intelligent optimization method driven by digital twins according to an embodiment of the present invention. Figure 1 As shown, the method includes:

[0027] S1. Construct a three-dimensional geometric model, material property model and physical property model of the transmission line, and establish a digital twin model of the transmission line by fusing multi-source heterogeneous data; wherein the three-dimensional geometric model includes geometric parameters of the tower, conductor, insulator string and hardware transmission line, the material property model includes the material model of the conductor, the material parameters of the dielectric constant and dissipation factor of the insulating material, and the physical property model includes the electrical parameters of the resistivity, inductance and capacitance of the conductor, and the physical parameters of the mechanical strength of the tower and insulator string; a power quality monitoring unit is arranged at the measuring point of the transmission line, the power quality monitoring unit includes a high-precision voltage transformer, a current transformer, a harmonic analyzer, and a temperature and humidity sensor, and the power quality monitoring unit uploads the power quality data collected and analyzed to the edge computing server in real time through the optical fiber private network; the edge computing server pre-processes the power quality data, evaluates the health status of the power quality of the transmission line through an intelligent algorithm, generates a power quality evaluation report and uploads it to the cloud server;

[0028] S2. The cloud server receives the power quality assessment report uploaded by the edge computing server. If the power quality assessment report shows that the power quality does not meet the standard, the digital twin model is called to simulate the power quality optimization process of the transmission line in the virtual space by adjusting the parameters of the digital twin model. On the basis of solving the power flow equations, a multi-objective optimization algorithm is used to search for the optimal transmission line operation parameter combination that meets the power quality standard, including the conductor cross-sectional area, reactive power compensation capacity, and ambient temperature and humidity control threshold;

[0029] S3. The optimal transmission line operating parameter combination obtained by optimizing the digital twin model is sent to the intelligent control system of the transmission line. The intelligent control system coordinates the reactive compensation device, the dynamic voltage regulator, and the intelligent de-icing device to dynamically adjust the reactive power, voltage amplitude and phase, and conductor temperature parameters of the transmission line so that the transmission line reaches the optimal operating state under the current environmental conditions and meets the power quality requirements. The intelligent control system feeds back the optimized control effect to the digital twin model for correcting and updating the relevant parameters of the digital twin model.

[0030] In an optional embodiment,

[0031] Collect 3D point cloud data of transmission line towers, conductors, insulator strings, and hardware, and perform denoising, registration, and fusion processing on the collected 3D point cloud data to improve the quality of 3D point cloud data; use a combination of parametric modeling and instance-based modeling to build a parametric 3D solid model, and generate 3D entities of different specifications through the instantiation method; use subdivision surface technology to perform personalized modeling for insulator strings and hardware; assemble the parametric 3D solid model into a complete 3D geometric model of the transmission line according to the spatial position relationship, and use scene tree and level of detail technology to optimize the organizational structure and rendering performance of the 3D geometric model;

[0032] According to the metal material grade of the conductor, the physical property parameters of the conductor material are obtained from the material database, including density, resistivity, temperature coefficient, and stress-strain curve, and a multi-physics field constitutive model of the conductor material is established; by testing the dielectric constant, dissipation factor, and breakdown voltage of the insulator string dielectric material, an insulator string dielectric material model that reflects the frequency and temperature dependence of the insulator string dielectric material is constructed; for the materials of the tower, foundation, and hardware, the corresponding material model is selected from the material database according to their material composition and mechanical performance indicators to construct a material property model for the transmission line;

[0033] On the basis of material property model and three-dimensional geometric model, the multi-physics field coupling model of the conductor is established by comprehensively considering the variation law of resistivity, inductance and capacitance of the conductor with temperature and stress; in view of the surface flashover and pollution flashover of the insulator string, the multi-physics field model of the insulator string is constructed, and the dielectric properties and discharge characteristics of the insulator string under different working conditions and environmental factors are quantitatively simulated; according to the material properties and structural characteristics of the tower, the mechanical model of the tower is established, and the finite element method is used to analyze the strength, stiffness and stability of the tower under wind load and ice load conditions; the multi-physics field coupling model of the conductor, the multi-physics field model of the insulator string and the mechanical model of the tower are integrated to construct the physical characteristic model of the transmission line.

[0034] The physical property model of the transmission line is used to construct a three-dimensional geometric model of the transmission line in a three-dimensional virtual environment, and the physical property model is matched with the three-dimensional geometric model through a mapping method to establish a digital twin model of the transmission line; neural networks and genetic algorithms are used to optimize the parameters of the digital twin model based on the historical operation data and online monitoring data of the transmission line, so that the simulation results of the digital twin model better fit the actual operating conditions; the coupling and association mechanism between the three-dimensional geometric model, material property model, and physical property model of the transmission line is explored, and a spatiotemporal synchronization mechanism and a data interaction interface between the three-dimensional geometric model, material property model, and physical property model are established to realize the fusion and coordination of multi-dimensional information in the digital twin model; the accuracy and reliability of the digital twin model are verified through theoretical analysis, experimental testing, and on-site measurement, the simulation results of the digital twin model are compared with the actual operation data, and the digital twin model is iteratively optimized.

[0035] Exemplarily, this paper proposes a multi-level, multi-dimensional digital twin model construction method for transmission lines. First, the three-dimensional point cloud data of key equipment of the transmission line is collected, and a high-precision three-dimensional geometric model of the equipment is generated through point cloud data preprocessing and parametric modeling; then, through material testing and database query, the material property parameters of the conductors, insulator strings, and towers are obtained to construct a multi-physics material constitutive model; based on the geometric model and material model, the multi-physics coupling characteristics of the transmission line are comprehensively analyzed, and a multi-physics simulation model of the conductors, insulator strings, and towers is established; finally, the geometric model, material model, and physical model are integrated and matched to form a digital twin model of the entire line scale, and the digital twin model is optimized and improved through parameter identification, multi-model fusion, performance verification and other means. The specific technical scheme is as follows:

[0036] 3D geometric modeling. Laser scanning, photogrammetry and other technologies are used to collect 3D point cloud data of transmission line towers, conductors, insulator strings and hardware. The collected raw point cloud data is pre-processed by denoising, registration, fusion and other methods to improve the accuracy and integrity of the point cloud data. Taking the 220kV transmission line as an example, the amount of tower point cloud data is about 15 million points, the point cloud density of the conductor is not less than 5 points / cm, and the point cloud density of the insulator string is not less than 10 points / cm. Based on the point cloud data, a strategy combining parametric modeling and instance-based modeling is used to build a 3D model of the equipment: for standardized equipment such as towers and conductors, key geometric parameters are extracted based on the design drawings, and a parametric 3D solid model is established. 3D entities of different specifications are quickly generated through the instantiation method; for non-standard equipment such as insulator strings and hardware, spline curves, subdivision surfaces and other technologies are used for personalized entity reconstruction. The equipment models are assembled into a complete 3D geometric model of the transmission line according to the spatial position relationship, and the data organization and rendering performance of the model are optimized by scene tree, frustum culling, and hierarchical details.

[0037] Material property modeling. Construct a multi-physics material constitutive model for key materials of transmission lines. For conductors, obtain the physical property parameters of metal materials such as density, resistivity, temperature coefficient, stress-strain curve from the material database, and establish a conductor material model of temperature-stress-electric-magnetic multi-field coupling. Taking the ACSR-720 / 50 conductor as an example, the resistivity is 0.05718Ω·mm² / m at 20℃, the temperature coefficient is 0.004 / ℃, and the density is 3.44×10³kg / m³. For insulator strings, a multi-physics insulator string material model that reflects the frequency and temperature characteristics of the dielectric material of the insulator string is constructed through test experiments on the dielectric constant, dissipation factor, and breakdown voltage. Taking the XWP2-160 insulator string as an example, the dielectric constant is 6.5, the dissipation factor is below 0.35%, and the breakdown voltage is not less than 160kV. For poles, towers, foundations, hardware and other equipment, the corresponding material models are selected from the material database according to their material grades and mechanical performance indicators. For example, the density of Q345 steel poles is 7850kg / m³, the elastic modulus is 206GPa, and the yield strength is 345MPa. Finally, a material property model that reflects the multi-field coupling characteristics of the transmission line is obtained.

[0038] Physical property modeling. Based on the geometric model and material model, the electric-magnetic-thermal-mechanical multi-physics coupling effect of the transmission line is comprehensively considered to construct the physical property model of the equipment. For the conductor, based on the mechanism of the change of resistivity, inductance and capacitance with temperature and stress, the electric-thermal-mechanical multi-physics model of the conductor is established to simulate the temperature rise curve and sag characteristics of the conductor under different current carrying levels and meteorological conditions. Taking the LGJ-400 conductor as an example, at an ambient temperature of 40℃, a wind speed of 1m / s and a current of 1000A, the steady-state temperature of the conductor is 70.5℃, and the corresponding sag is 6.72m. For the insulator string, for typical faults such as surface flashover and pollution flashover, the electric-thermal-humidity multi-physics model of the insulator string is constructed to quantitatively analyze the hot spot temperature and leakage current of the insulator string under different pollution levels and humidity conditions. Taking the FC160P / 146 insulator string as an example, when the salt density is r=0.1mg / cm², its breakdown voltage drops to 46.2kV. For the tower, the finite element method is used to establish the geometry-material-load multi-field model of the tower to simulate the deformation and stress distribution of the tower under wind deflection, icing and other working conditions. Taking the corner tower as an example, under the conditions of 10m / s wind speed and 10mm rime, the maximum displacement of its top is 0.52m and the maximum stress of the main material is 165MPa. Finally, a physical characteristic model reflecting the multi-physical field coupling mechanism of the transmission line is formed.

[0039] Twin model integration. Based on the above work, the three-dimensional geometric model, material property model, and physical property model of the transmission line are integrated and matched to construct a digital twin model of the entire line scale. The semantic mapping method is used to associate the material properties, boundary conditions and other information in the physical model with the geometric elements such as points, lines, and surfaces in the geometric model to achieve synchronization between physical properties and geometric entities. Using data-driven neural networks, genetic algorithms and other intelligent optimization technologies, based on the historical operation data and online monitoring data of the transmission line, the key parameters in the digital twin model are identified and corrected to improve the fitting accuracy of the simulation model to the actual system. Taking the 220kV transmission line as an example, the prediction error of the tower displacement can be reduced to less than 5% through optimization. In-depth analysis of the spatiotemporal evolution laws and data association mechanisms of the geometric, material, and physical three-dimensional models, design data interaction interfaces and synchronous update mechanisms between models, and realize a full-dimensional, strongly coupled digital twin model. The accuracy and reliability of the digital twin model are evaluated through theoretical analysis, experimental testing and field verification. Based on the deviation between the simulation results and the measured data, the parameters and structure of the twin model are iteratively optimized to construct a high-fidelity and reliable digital twin model of the transmission line.

[0040] In summary, this paper proposes a multi-level and multi-dimensional digital twin modeling method for power transmission lines, covering the entire process from data collection, model construction to integrated simulation. By acquiring point cloud data through 3D laser scanning and performing parametric modeling, a high-precision geometric model was constructed; a multi-field coupled material property model was established in combination with material testing and database analysis; an electric-magnetic-thermal-mechanical multi-physics field model was formed for the physical mechanisms of conductors, insulator strings, and towers; mapping matching, parameter identification, performance verification and other methods were used to achieve seamless integration and dynamic optimization of multi-dimensional models, and finally a high-fidelity digital twin model of the entire line scale was obtained. The constructed digital twin model lays the foundation for the precise design, intelligent operation and maintenance, situation warning and other applications of power transmission lines.

[0041] In an optional embodiment,

[0042] The cloud server receives the power quality assessment report uploaded by the edge computing server. If the power quality assessment report shows that the power quality does not meet the standard, the digital twin model is called, and the process of optimizing the power quality of the transmission line is simulated in the virtual space by adjusting the parameters of the digital twin model. On the basis of solving the power flow equation group, the multi-objective optimization algorithm is used to search for the optimal combination of transmission line operating parameters that meet the power quality standard, including the following steps:

[0043] Exemplarily, this paper proposes a power quality optimization control strategy for transmission lines based on digital twins, which uses a cloud-edge-end collaborative architecture to achieve bidirectional mapping and closed-loop optimization between digital twin models and physical transmission lines. The edge computing server monitors the physical transmission lines in real time, evaluates the power quality level, and generates a power quality assessment report and uploads it to the cloud; after the cloud server receives the power quality assessment report, if the power quality does not meet the standard, the digital twin model is called to optimize the line operation parameters based on multi-objective optimization, and the optimization results are decoded into control instructions and sent to the transmission line intelligent terminal to achieve digital twin-driven physical line optimization control. The specific technical solution is as follows:

[0044] Multi-scenario and multi-objective optimization modeling. In view of the excessive power quality received in the cloud, a multi-scenario and multi-objective transmission line optimization control problem is constructed in the digital twin model. Taking the voltage qualification rate, harmonic content, and network loss rate as the optimization targets, and considering the typical operation scenarios of different load levels and new energy output, the line impedance, interconnection switch status, and tap position of the digital twin model are extracted to form constraints. Taking a 500kV transmission line as an example, its optimization target is: the voltage qualification rate is not less than 99.5%, the total harmonic distortion rate does not exceed 3%, and the network loss rate does not exceed 2%; the constraints include: the line flow does not exceed 90% of the thermal stability limit, the node voltage amplitude is maintained at 0.95-1.05pu, the reactive compensation capacity adjustment range is 0-200MVar, and the tap adjustment range is ±8×1.25%. On this basis, combined with load forecasting, new energy power forecasting and other data, four typical operation scenarios such as high load, low load, high wind power output, and high photovoltaic output are designed to form a multi-objective optimization model under multi-scenario constraints.

[0045] Power flow calculation and multi-objective optimization solution. The network topology, branch parameters, bus type and other information contained in the digital twin model are encoded in standard formats such as IEEE CDF to automatically generate a group of power flow equations. The fast decoupling method is used to solve the power flow equations under different operating parameters to obtain the voltage amplitude, phase angle, branch active and reactive power flow distribution of the digital twin model under different working conditions. On this basis, the improved multi-objective evolutionary algorithm MOEA / D is used to solve the multi-scenario multi-objective optimization model. In terms of algorithm performance, the topological structure, physical mechanism and other problem knowledge of the transmission line are incorporated into the evolution process, and targeted improvements are made to the population initialization and crossover mutation operators, such as using topological neighborhoods to select parent individuals, adaptively controlling the crossover mutation probability, etc., and using external archiving strategies, adaptive grid division and other methods to ensure the convergence and distribution of the solution, and finally search for multiple dimensionless Pareto optimal solutions. Taking a 220kV line as an example, 25 optimal solutions can be obtained through 1000 generations of evolution, among which the voltage qualification rate is increased from 98.5% to 99.8%, the harmonic content is reduced from 4.2% to 2.5%, and the network loss rate is reduced from 1.8% to 1.3%.

[0046] Parameter decoding and digital twin mapping. The Pareto solution set obtained by multi-objective optimization is analyzed. By constructing a parameter sensitivity matrix, the weight of each parameter's influence on different objectives is quantified, and the key parameters that contribute most to the improvement of power quality are screened out. These parameters are adjusted in a targeted manner in the digital twin model, and the optimized power quality level is verified through full-scale simulation. The impact of parameter adjustment on equipment safety and network stability is evaluated, and finally a set of optimal parameter combinations that can achieve power quality standards under multi-scenario constraints are determined. Taking the improvement of voltage qualification rate as an example, through analysis, it is found that the voltage amplitude can be controlled within 0.97-1.03pu under different load levels by reasonably configuring the reactive compensation capacity of 10 nodes, so that the voltage qualification rate can be stabilized at more than 99.5%. The optimal parameter combination output by the digital twin model is converted into control instructions through a standard interface, including the SVG switching power / number of units, OLTC tap position, generator stator terminal voltage set value, etc. The control instructions are decoded and sent to the intelligent operation and maintenance terminal using the edge computing server, and the reactive compensation device, transformer tap, generator excitation and other equipment of the physical transmission line are coordinated and controlled to achieve closed-loop optimization control driven by the digital twin.

[0047] Closed-loop measurement and control integration. The digital twin system achieves data interoperability and process collaboration with the monitoring system and dispatching control system of the physical transmission line through a standard interface, and builds a full-process closed-loop measurement and control system. During the operation of the physical transmission line, intelligent sensors, PMU and other monitoring equipment continuously collect voltage, current and other indicator data and upload them to the digital twin model, and calibrate the digital twin model in real time to synchronize its state with the physical line; the optimization control instructions output by the digital twin model are executed through the intelligent control terminal, and the power quality level of the physical line is improved. The optimization control effect is fed back through the monitoring equipment to evaluate the reliability and effectiveness of the optimization instructions; when the monitoring system finds that the power quality fluctuates greatly and there is a risk of exceeding the limit, the digital twin model is triggered through cloud-edge collaboration to carry out early warning analysis and dynamic optimization, thereby improving the adaptability and real-time performance of the digital twin. Through the closed-loop measurement and control, the dynamic mapping, interactive optimization and collaborative evolution of the digital twin model and the physical line are realized.

[0048] In summary, this paper proposes a power quality optimization control strategy for transmission lines based on digital twins. Based on the cloud-edge-end collaborative architecture, it realizes the real-time mapping of physical line monitoring data to digital twin models and the closed-loop control of digital twin model optimization results to physical lines. Through multi-scenario and multi-objective modeling, knowledge-driven evolutionary optimization algorithm, parameter decoding and full model simulation, the optimal control parameter combination that meets the power quality standards is searched in the digital twin model; the interaction between the digital twin model and the physical line is realized by using standardized interfaces, and the line voltage qualification rate is improved, the harmonic content and network loss rate are reduced by regulating the reactive compensation, transformer tap, generator excitation and other equipment of the physical line; at the same time, a full-process measurement and control closed loop is constructed to ensure the accuracy and reliability of the digital twin model, providing a new solution for the full life cycle management of power quality of transmission lines.

[0049] In an optional embodiment,

[0050] In response to the situation where the power quality exceeds the standard, a multi-scenario and multi-objective optimization control problem is constructed, and the optimization objectives of power quality indicators, transmission line thermal capacity margin, and network loss are comprehensively considered. In addition, the load forecast and new energy output forecast data are combined to generate the constraints under multiple operation scenarios. The steps include:

[0051] For example, in response to the problem of excessive power quality in transmission lines, this paper proposes a multi-scenario and multi-objective optimization control strategy. By constructing a weighted multi-objective function that comprehensively considers power quality, line safety and economic operation, establishing multi-dimensional constraints, generating multi-scenario operation data sets, and using intelligent algorithms to solve the optimization model, the optimal control strategy that meets the multi-scenario constraints is obtained, and real-time closed-loop control of the transmission line is achieved through the distribution automation system. The specific technical solution is as follows:

[0052] Construct a multi-objective optimization function. According to the power quality assessment report, extract key indicators such as voltage deviation, harmonic content, line load rate, network loss level, etc., and construct corresponding sub-objective functions. Among them, the voltage deviation is measured by the sum of the squares of the difference between the voltage amplitude of each bus and the nominal value, the harmonic content is measured by the sum of the squares of the difference between the voltage / current harmonic content of each bus and the limit value, the line current margin is measured by the sum of the squares of the difference between the current of each line and the thermal stability limit, the network loss is measured by the ratio of the active loss of the whole network to the total load, and the reactive power optimization amount is measured by the total reactive power flow of the line. On this basis, the weight coefficient is introduced to perform weighted summation on each sub-objective to form a multi-objective optimization function. Taking a 220kV transmission line as an example, its multi-objective function contains 5 sub-objectives, and the weight coefficients are 0.3, 0.25, 0.2, 0.15, and 0.1, respectively, reflecting the higher priority of power quality and line safety over network loss saving.

[0053] Establish multi-dimensional constraints. Considering the physical model and safe operation requirements of the transmission line, construct multi-dimensional constraints. The equality constraint consists of the active and reactive power balance equations of the bus node, reflecting the physical mechanism of the power flow. The inequality constraint includes the bus voltage amplitude constraint and the line current constraint, where the bus voltage must be maintained within ±5% of the nominal value and the line current must not exceed 90% of the thermal stability limit. The control variable constraint consists of the power quality index limit and the physical adjustment range of the reactive compensation device, transformer tap, and generator excitation. Taking a 500kV line as an example, the equality constraint consists of 250 power balance equations, the inequality constraint includes 500 voltage amplitude constraints and 100 line current constraints, and the control variable constraint involves the reactive output of 20 SVCs, 10 OLTC transformers, and 5 generators.

[0054] Generate multi-scenario operation data sets. Use load classification and clustering technology to mine the typical daily change pattern of load based on historical data, and consider the influence of factors such as seasons, weather, holidays, etc., predict the range of load changes in the future, and divide the load levels into peak, flat, and valley levels. Introduce new energy output prediction models such as random forests and support vector machines, comprehensively consider the influencing factors of wind power and photovoltaic power generation, such as wind speed, irradiance, temperature, etc., and generate new energy output levels and their probability distributions under multiple scenarios. On this basis, through Monte Carlo simulation, random sampling is used to form multiple combinations of transmission line impedance parameters, tie switch switching states, etc., which are matched with load and new energy output scenarios, and finally a multi-scenario data set that fully covers typical operating conditions is obtained. Taking a 220kV line as an example, four typical load curves are obtained through clustering, and 10 line parameter combinations are obtained by random sampling, generating 40 operation scenarios, including 15 high-load and heavy-load scenarios, 10 low-load and light-load scenarios, 8 high wind and solar penetration scenarios, and 7 extreme N-1 scenarios.

[0055] Solving the multi-objective optimization model. The multi-objective evolutionary algorithm is used to solve the multi-scenario and multi-objective transmission line optimization model to obtain the optimal control strategy that meets the power quality compliance constraints. The algorithm flow is as follows: First, the initial population is randomly generated, and the individuals are composed of control variables such as reactive compensation capacity, tap position, and generator terminal voltage setting value. For each individual, the power flow calculation program is called in turn to analyze its electrical parameters under all operating scenarios, and the state variables are substituted into the objective function and constraints. The constraint violation is quantified using the static penalty function method, and the multi-objective optimization problem is converted into a single-objective optimization problem by weighted summation. Then, iteratively execute evolutionary operators such as selection, crossover, and mutation to update the fitness and violation of the population, retain non-dominated solutions, and eliminate inferior solutions. Until the maximum evolutionary generation or performance index requirements are met, the Pareto optimal solution set is output. Taking a 500kV line as an example, 25 optimal solutions were obtained after 2000 generations of evolution, among which the voltage qualification rate was increased from the original 97% to more than 99.5%, the harmonic distortion rate was reduced from 4.5% to less than 3%, the line current margin was increased from 5% to 15%, and the network loss rate was reduced from 2.2% to 1.8%.

[0056] Analysis, evaluation and automated control. Conduct adaptability and economic analysis on the optimization results, evaluate the effect of improving power quality under different operating scenarios, and the resulting network loss savings and equipment investment savings. Use Monte Carlo simulation and other methods to analyze the ability of the optimization control strategy to resist random disturbances such as load fluctuations and new energy output, and evaluate its robustness and reliability. Combined with the transmission line automation system and energy management system, the optimization results are converted into executable remote control instructions, such as adjusting SVC switching, OLTC tap raising and lowering, generator excitation regulation, etc., and the monitoring system is used for real-time measurement and feedback to form an optimization control closed loop. At the same time, if the line conditions change significantly beyond the scope of the optimization scenario, the real-time correction of the line digital twin model is triggered, and new multi-scenario constraints consistent with the physical line mapping are generated. The optimization calculation is re-conducted so that the optimization control strategy can adapt to the dynamic changes in external conditions.

[0057] In summary, this paper proposes a multi-scenario and multi-objective transmission line optimization control strategy, which is elaborated in detail from the aspects of constructing weighted multi-objective functions, establishing multi-dimensional constraints, generating multi-scenario operation data, solving optimization models with intelligent algorithms, analyzing and evaluating optimization results, etc. By placing power quality, line safety and economic operation in the same optimization framework, using big data analysis and machine learning to improve the accuracy and comprehensiveness of the optimization model, using the global optimization capability of modern heuristic intelligent algorithms to obtain the optimal control strategy that meets multi-scenario constraints, and using distribution automation and energy management systems to implement the engineering implementation of optimization control, it provides new ideas and methods for ensuring the power quality of transmission lines in complex scenarios. In the next step, further research is planned on the digital twin of transmission lines, the integration of physical information and business processes, etc., to lay the foundation for building an intelligent optimization control system for the entire life cycle of transmission lines.

[0058] In an optional embodiment,

[0059] The cloud server receives the power quality assessment report uploaded by the edge computing server. If the power quality assessment report shows that the power quality does not meet the standard, the digital twin model is called to simulate the process of power quality optimization of the transmission line in the virtual space by adjusting the parameters of the digital twin model. On the basis of solving the power flow equation group, a multi-objective optimization algorithm is used to search for the optimal transmission line operation parameter combination that meets the power quality standard, including the conductor cross-sectional area, reactive power compensation capacity, and ambient temperature and humidity control threshold. The steps include:

[0060] For example, for the power quality assessment report uploaded by the edge computing server, this paper proposes a transmission line power quality optimization method based on digital twin and multi-objective optimization. After the cloud server receives the assessment report that the power quality does not meet the standard, it calls the digital twin model of the transmission line to simulate the optimization process in the virtual space. By adjusting parameters such as the line conductor cross-sectional area, reactive compensation capacity, and ambient temperature and humidity control threshold, an improved multi-objective particle swarm optimization algorithm is used to search for the optimal parameter combination that meets multiple power quality indicators, thereby guiding the intelligent optimization control of the transmission line. The specific technical solution is as follows:

[0061] Particle encoding and initialization. A hybrid encoding strategy is used to encode the control variables of the transmission line. Continuous variables such as reactive compensation capacity are encoded with real numbers, and discrete variables such as conductor type are encoded with binary numbers. Each particle represents a combination of transmission line parameters, and the particle dimension is equal to the number of control variables. The initial particle swarm is randomly generated, and the power quality indicators such as voltage deviation, harmonic distortion rate, power factor, etc. of each particle in a typical operation scenario are calculated using the power flow calculation module. Each indicator is dimensionless, and a weighted fitness function is constructed as a particle performance evaluation indicator. Taking a 220kV line as an example, each particle contains four control variables: conductor type, reactive compensation capacity, temperature control upper limit, and humidity control upper limit. There are four options for conductor type, which are encoded as 00 / 01 / 10 / 11. The remaining variables are encoded with real numbers in the range of [0, 50] MVar, [30, 50] ℃, and [60, 80]%. 100 particles are generated as the initial population, and the weight vector is set to [0.5, 0.3, 0.2], indicating that the voltage qualification rate has the highest priority, followed by the harmonic compliance rate and the power factor margin.

[0062] Particle selection and crossover. The initial particle swarm is sorted non-dominatedly according to the fitness function value, and the particles are stratified according to their Pareto levels. The level of the particles at the Pareto front is 1, and the levels of the remaining particles are increasing in sequence. The crowding distance of the particles in the same layer is calculated. The larger the distance, the higher the solution sparsity in the area where the particles are located. Based on the particle level and crowding distance information, the binary tournament selection operator is used to select the parent particle from the current particle swarm. In the crossover operation, the power quality index weight information is introduced to adaptively adjust the crossover probability. The larger the weight, the higher the control variable crossover probability. At the same time, the infeasible solution generated by the crossover is repaired using the discrete variable boundary constraint rule and the continuous variable calibration mapping rule to ensure the feasibility of the child particles. Taking the above 220kV line as an example, the particles are divided into 8 levels through non-dominated sorting. The sum of the crowding distances of the first 3 layers of particles accounts for 65%, and 50 particles are selected as the parent. The crossover probability vector is set to [0.8, 0.6, 0.4, 0.2], that is, the conductor type has the highest probability of crossover, followed by reactive compensation capacity, temperature upper limit and humidity upper limit.

[0063] Particle mutation and merging. The sensitivity information of power quality to the control variables is introduced to adaptively adjust the particle mutation probability and the step length. The higher the sensitivity of the variable, the greater the mutation probability and the smaller the step length. Conversely, the mutation probability is small and the step length is large. The mutation operation perturbs the control variables of the parent particles to generate diverse child particles. The parent and child particles are merged to form a new particle group, and the merged particle group is re-sorted in a non-dominated manner. Taking a 500kV line as an example, the sensitivity analysis shows that the harmonic distortion rate is most sensitive to the reactive compensation capacity, followed by the conductor cross-sectional area, and the temperature and humidity thresholds have little effect. Based on this, the adaptive mutation probability vector is set to [0.6, 0.8, 0.2, 0.2] and the step length vector is set to [2, 5, 10, 10]%. After mutation, the number of child particles is the same as that of the parent, and the size of the particle group after merging is expanded to 200. The merged particle group is re-sorted, the number of dominated solutions increases to 35, and the population diversity is improved.

[0064] Particle selection and external archiving. Based on the Pareto level and crowding distance information of particles, an elite retention strategy is used to select a new generation of particle swarms from the merged particle swarm. Particles with smaller levels and larger crowding distances are retained first until the number of new particle swarms reaches the preset size. At the same time, the non-dominated solutions of the current particle swarm are retained in the external archive set to participate in the fitness evaluation of the next generation of particles, and the weights of each indicator are dynamically adjusted according to the degree of satisfaction of the power quality indicators. The weights of indicators with high satisfaction levels decrease, guiding particles to search in the direction of under-satisfied indicators. Taking a 220kV line as an example, 100 particles are selected from the merged particle swarm to form a new generation of populations, among which the proportion of particles with levels 1 and 2 is increased to 60%. 25 non-dominated solutions are stored in the external archive set, and the weight vector is updated to [0.3, 0.2, 0.5] according to the compliance rates of the three indicators (such as 95%, 98%, and 90%), which improves the optimization weight of the power factor.

[0065] Constraint processing and iterative update. The adaptive penalty function method is used to deal with constraint violation problems for the new generation of particle swarms. For control variables that exceed the physical boundary, the particle fitness is penalized according to the degree of violation. The greater the degree of violation, the greater the penalty intensity. The fitness function after the penalty guides the particles to search within the feasible domain. In addition, the standard particle swarm algorithm is used to update the speed and position of continuous variables, and the quantum behavior strategy is introduced to update the revolving door for discrete variables to balance the local development and global exploration capabilities of the algorithm. Taking a 500kV line as an example, 5% of the individuals in the new generation of particle swarms violate the upper limit constraint of the reactive compensation capacity. According to the amplitude of the excess, their fitness values ​​are reduced by 20%~50%. The random weight strategy is used to iterate the speed and position of the reactive compensation capacity and temperature and humidity thresholds in the continuous variables, and the learning factor is set to 2.0. For the conductor type variable, the rotation angle of the quantum revolving door is adaptively adjusted by measuring the Hamming distance between the particle and the optimal solution, and the angle range is set to [0.2π, 0.8π].

[0066] Termination conditions and result output. Repeat the iterative evolution of the particle swarm in steps 2 to 5 until the maximum evolutionary generation or particle swarm convergence requirements are met. The maximum evolutionary generation is set to 300. When the optimal fitness value increases by no more than 0.1% for 50 consecutive generations, the particle swarm is considered to converge. Output all particles in the external archive set as the Pareto optimal solution set for the multi-scenario transmission line optimization control problem. Each optimal solution corresponds to a parameter combination of conductor cross-sectional area, reactive compensation capacity, and temperature and humidity thresholds. Under the premise of meeting multiple power quality index constraints, it weighs the performance requirements of line safety and economy. Taking a 220kV line as an example, 12 optimal solutions were obtained after 300 generations of evolution, among which the voltage qualification rate exceeded 99%, the harmonic distortion rate was less than 3%, and the power factor was above 0.9. The average conductor cross-sectional area of ​​the optimal solution set was reduced by 8% compared with the initial value, the average reactive compensation capacity was reduced by 12%, and the adjustment range of the temperature and humidity threshold did not exceed 5%, which significantly reduced the line loss and investment cost while ensuring that the power quality met the standards.

[0067] In summary, this paper proposes a method for optimizing the power quality of transmission lines by calling digital twin models on cloud servers. This method receives the power quality assessment report uploaded by the edge computing server. When the power quality exceeds the standard, the optimization control process is simulated in the virtual space by adjusting the parameters of the digital twin model. The hybrid coded multi-objective particle swarm optimization algorithm is used to search for the optimal parameter combination of the transmission line that balances multiple power quality indicators based on the power flow calculation. The algorithm's optimization ability and convergence speed are enhanced by integrating the crossover operator, adaptive mutation strategy, and elite retention mechanism with domain knowledge. At the same time, the penalty function method is introduced to deal with constraint violation problems, and the quantum behavior strategy is used to update discrete variables, which improves the algorithm's ability to deal with complex constraints and mixed variables. The output results of this method are a set of transmission line optimization control strategies under multiple scenarios, which can be used as a reference for dispatchers and executed by the automation system, providing a new solution for ensuring that the power quality meets the standard and improving the comprehensive performance of the transmission line. In the next step, it is planned to further improve the accuracy and real-time performance of the digital twin model, expand the external scenarios applicable to the optimization algorithm, and combine intelligent optimization technologies such as reinforcement learning to improve the intelligence level of power quality control of transmission lines.

[0068] In an optional embodiment,

[0069] The optimal transmission line operation parameter combination obtained by optimizing the digital twin model is sent to the intelligent control system of the transmission line. The intelligent control system collaboratively controls the reactive compensation device, the dynamic voltage regulator, and the intelligent deicing device to dynamically adjust the reactive power, voltage amplitude and phase, and conductor temperature parameters of the transmission line so that the transmission line reaches the optimal operating state under the current environmental conditions and meets the power quality requirements. The steps include:

[0070] For example, this paper proposes a multi-level collaborative intelligent optimization control method for transmission lines based on the optimal control parameters of the digital twin model optimization output. This method realizes the closed-loop interaction of digital twins, intelligent control and physical equipment from the three levels of master station layer, substation layer and equipment layer. Through the coordinated regulation of multiple devices, the reactive power, voltage parameters and conductor temperature of the transmission line are dynamically optimized to ensure that the power quality of the line meets the standards and the comprehensive benefits are maximized under the current environmental conditions. The specific technical scheme is as follows:

[0071] Generation and issuance of optimized control instructions. The digital twin model outputs a set of optimal transmission line operation parameter combinations for the current external environment, including reactive power, voltage amplitude, voltage phase and conductor temperature. Taking a 500kV line as an example, when the ambient temperature is 25℃, the humidity is 60%, and the wind speed is 3m / s, the optimal parameters given by the digital twin model are: reactive power 30MVar, voltage amplitude 525kV, voltage phase 5°, and conductor temperature 45℃. The master station layer transmits the above parameters to the smart substation terminal (DTU) and the line local controller (LCU) through the optimized control instruction issuance interface. DTU and LCU divide the control domain according to the substation and line segment connected to the line, and each control domain contains multiple types of controllable devices. After receiving the optimized control instruction, the DTU decomposes the total instruction into multiple device-level executable control strategies according to the topological relationship of the equipment in the station, and forwards it to the corresponding intelligent switch device. LCU directly converts the optimized control instruction into the control strategy of this section of the line.

[0072] Formulate a multi-device collaborative control plan at the substation level. The substation-level intelligent controller (DTU / LCU) comprehensively considers the state parameters of each controllable device within the jurisdiction, including adjustment range, adjustment accuracy, adjustment speed, reliability, etc., according to the control strategy, and calculates the comprehensive benefit of each device through a multi-device benefit evaluation model. Taking a certain smart substation as an example, the transformer tap adjustment range is ±8×1.25%, the capacitor bank capacity is 100MVA, 40 taps are switched, the action time is 1s, the SVG rated capacity is 50Mvar, and the response time is 10ms. Based on the equipment parameters, the DTU constructs a benefit matrix and uses the fuzzy multi-attribute decision method to calculate the comprehensive benefit of the equipment. The results show that the benefits of the transformer, capacitor bank, and SVG are 0.8, 0.6, and 0.9, respectively. Based on this, the DTU formulates a multi-device collaborative control plan of "prioritized SVG action, capacitor bank assistance, and transformer tap fine-tuning", which is converted into executable instructions for controlling each device.

[0073] Primary equipment executes optimized control in parallel. The substation layer sends the executable instructions of the equipment to the primary equipment, including smart transformers, static VAR compensators (SVC / SVG), dynamic voltage restorers (DVR), and intelligent deicing devices for transmission lines. The equipment executes control instructions in parallel, and coordinates the reactive power, voltage parameters, and conductor temperature of the line by adjusting the transformer tap position, switching the capacitor bank tap, changing the SVG output reactive power, and optimizing the deicing device operating parameters. Taking a certain smart substation as an example, the transformer tap is adjusted to the 5th gear, 20 capacitor banks are put into use, and the SVG outputs 35MVar capacitive reactive power. In conjunction with the control of adjacent substations, the reactive power of the line is controlled at about 30MVar, the voltage amplitude is maintained at 525kV, and the phase angle is increased to 5°. The intelligent deicing device for transmission lines controls the conductor temperature within 45°C by optimizing the deicing current size and power-on time, avoiding the problem of corona and increased line loss caused by overheating of the line.

[0074] Equipment operation status perception and data fusion. Intelligent sensors at the equipment layer collect line voltage, current, power, frequency and other parameters as well as the equipment's own operating conditions in real time. Taking a 220kV line as an example, 6 sets of merged unit intelligent monitoring devices are arranged, with a collection frequency of 0.5s / time and a measurement accuracy of 0.2. The substation bus voltage reaches 220kV, the line apparent power is 200MVA, and the power factor is 0.95. At the same time, the state quantities such as the transformer oil level and the SVG DC side voltage are all within the qualified range. The substation layer aggregates the equipment layer data through the local optical fiber network and wireless MESH network, and after time-space synchronization, data compression, abnormal diagnosis and other processing by the edge computing node, it is uploaded to the digital twin model through the equipment operation status synchronization interface and power quality data upload interface of the main station layer for real-time correction and power quality evaluation.

[0075] Power quality analysis, early warning and closed-loop control optimization. The power quality management system is deployed at the main station layer to continuously monitor power quality indicators such as voltage deviation, harmonic distortion rate, and three-phase imbalance of the line. Once the power quality exceeds the standard, the system automatically triggers the digital twin model to perform optimization again, and prompts the line abnormality through the intelligent alarm module. Taking a 500kV line as an example, when the voltage qualification rate drops below 95%, the digital twin model combines external source data such as real-time meteorological conditions, load power forecasts, and new energy output forecasts to quickly iterate and optimize, and give a new optimal line operation parameter combination within 2 minutes. The reactive power is increased to 35MVar, the voltage amplitude is 530kV, and the temperature upper limit is 48℃. The closed-loop feedback control guides the dynamic adjustment of the line. In addition, line fault information and equipment health information are also fed back in time through the data link. The digital twin model periodically rolls and optimizes, outputs new control instructions, and realizes continuous optimal operation of the entire life cycle of the transmission line through real-virtual mapping and closed-loop interaction.

[0076] In summary, this paper proposes an intelligent optimization control method for transmission lines based on digital twins. This method receives the optimal control parameters output by the digital twin model and carries out multi-level collaborative control from the master station layer, substation layer, and equipment layer. The master station layer is responsible for parsing and issuing control instructions, the substation layer formulates a multi-device collaborative control plan, and the equipment layer is responsible for physical execution optimization control. Intelligent sensors sense the line status and equipment conditions in real time, and complete data aggregation and upload through the communication network. The power quality management system continuously evaluates the power quality level of the line, triggers the re-optimization of the digital twin model when an abnormality occurs, and realizes the continuous optimal control of the transmission line driven by the digital twin through equipment status mapping and closed-loop feedback of optimization results, ensuring that the power quality of the line meets the standard and the comprehensive benefits are maximized. The next research focus is to further simplify the optimization control model, improve the real-time and interpretability of the control instructions, explore the panoramic collaborative optimization control method of the line combined with digital twins and intelligent operation and inspection, and build an integrated simulation verification platform for digital twin models and line physical equipment to promote the application of the method in actual engineering.

[0077] In an optional embodiment, the method further includes:

[0078] For example, based on the digital twin optimization control, this paper further explores the coordinated optimization control strategy of voltage and icing temperature for two typical working conditions of transmission line voltage deviation and icing. Through the anomaly detection and multi-time and space scale simulation prediction of the digital twin model, the risk hazards of line voltage over-limit and icing temperature rise are accurately identified, and the intelligent control process is actively triggered. In terms of voltage optimization control, a hierarchical multi-station multi-equipment collaborative adjustment method is adopted. Through the combined switching of transformer taps, reactors and capacitor groups, the voltage of key nodes of the line is accurately adjusted to ensure that the voltage deviation is within the safety threshold. In terms of icing optimization control, differentiated control areas are divided according to the shadow position and boundary of the icing section. Through intelligent deicing in different zones, the operating parameters of the deicing device are accurately adjusted to ensure that the icing temperature meets the requirements while reducing energy consumption and costs. Both optimization controls adopt bottom-up data fusion and top-down command transmission methods to construct full-cycle closed-loop control to realize active prevention and control of transmission line voltage and icing temperature driven by digital twins. The specific technical solutions are as follows:

[0079] Intelligent identification of voltage anomalies and coordinated regulation of multiple stations. The digital twin model continuously monitors the real-time voltage of each measuring point of the line to determine the extent to which the voltage deviates from the optimal value. Once the voltage deviation exceeds the specified threshold (such as ±5%), the voltage optimization control is triggered. Taking a 500kV line as an example, the digital twin model sets the voltage qualification range to 525kV±5%. When it is detected that the bus voltages of substations A, B, and C are 512kV, 530kV, and 508kV respectively, and the average line voltage is 516kV, the deviation from the optimal voltage of 525kV is 1.7%, and the voltages of substations A and C have exceeded the lower limit, the model automatically switches to the voltage optimization control mode. Based on the three-phase AC optimal power flow (OPF) algorithm, the digital twin model takes the minimum comprehensive cost of power grid operation as the objective function, and quickly calculates the bus voltage regulation of each substation through iterative optimization. The calculation results show that the voltage regulation of substations A, B, and C are +6kV, -3kV, and +8kV respectively. The adjusted bus voltages are 518kV, 527kV, and 516kV, the average line voltage reaches 520kV, and the voltage qualification rate is increased to 100%. The digital twin model generates the corresponding voltage regulation control instructions, which are transmitted to the master station layer through the control instruction issuing interface.

[0080] Optimal voltage regulation solution analysis for station-side intelligent terminals. The master station layer is responsible for receiving and parsing the voltage optimization control instructions of the digital twin model, extracting the total amount of regulation of each substation intelligent terminal (DTU), and forwarding the decomposed instructions to each DTU through the optical fiber communication network. The DTU further refines the list of equipment that can participate in voltage regulation in this station and its comprehensive benefits, mainly including transformer taps, reactors, capacitors and other reactive compensation devices. Based on the "benefit-cost" of the regulation action of a single device, the DTU uses the fuzzy multi-attribute decision method to assign benefit weights and cost weights to each regulation action. The benefit weight mainly considers indicators such as voltage regulation effect and regulation speed response, and the cost weight mainly considers indicators such as equipment loss and investment and operation and maintenance costs. The DTU calculates the comprehensive benefits of each device, searches for a set of multi-device combination regulation solutions with the goal of maximizing benefits, and selects the optimal voltage regulation solution through the decision tree algorithm. Taking substation A as an example, the equipment that can participate in the regulation includes 2 transformers, 1 group of reactors and 4 groups of capacitors. The benefit weights of each equipment are 0.8, 0.6 and 0.4 respectively, and the cost weights are 0.2, 0.4 and 0.6 respectively. Based on this, the comprehensive benefits of a single device action can be calculated. Through the combination of equipment, DTU searches for 5 optional regulation schemes, and through multi-attribute decision-making, determines the optimal voltage control scheme of "prioritizely adjusting transformer No. 1 to the 8th gear, putting in 1 group of reactors, and removing 2 groups of capacitors". DTU sends the executable control instructions of each device in the optimal scheme to the local intelligent controller.

[0081] Parallel control at the equipment level achieves precise voltage regulation. After receiving the control instructions of this device, the intelligent controller (LCU) at the equipment level converts them into executable code to drive the actuator action. In the transformer LCU, the control logic converts the "adjust to the 8th gear" instruction into "the current position is the 6th gear, and 2 more gears are needed", driving the electric mechanism of the transformer on-load tap changer to switch the tap from the 6th gear to the 8th gear, and the transformer secondary voltage increases by 2×1.25%. In the reactor and capacitor LCU, the actions of putting in one group of reactors and removing two groups of capacitors are respectively executed. Under the unified dispatch of the DTU, the control actions of the transformer, reactor, and capacitor are carried out synchronously, which takes 2 minutes and 30 seconds to complete, and the bus voltage of substation A is adjusted from 512kV to 520kV. Similarly, the DTUs of substations B and C decompose the station adjustment amount according to the instructions of the master station, formulate the optimal combination adjustment plan of transformers, reactors, and capacitors, and control related equipment in parallel, and finally adjust the bus voltage to 527kV and 516kV. The voltages at key nodes of the transmission lines have all returned to the qualified range specified by the digital twin model, and the average voltage of the lines has been increased to 520kV, with the voltage qualification rate increased from 83% to 100%.

[0082] Icing section division and differentiated intelligent deicing. The transmission line icing monitoring device continuously collects meteorological conditions and conductor temperature, and the digital twin model analyzes the growth trend of conductor icing. When it is predicted that the icing temperature will exceed the warning value (such as 5°C) in the next 2 hours, the digital twin system promptly sends deicing control instructions to the intelligent control system. The instructions contain icing risk section information (such as the section from G25# tower to G38# tower) and the target deicing temperature (such as -3°C). The intelligent control system parses the deicing instructions, combines high-precision meteorological radar and satellite remote sensing analysis, and accurately tracks the movement trajectory of low-temperature clouds that affect icing. According to the projection position of the cloud shadow on the line icing section, the section is divided into three types of control sub-areas: shadow center (such as G28# to G32#), shadow transition area (G25# to G27#, G33# to G35#), and shadow edge (G36# to G38#), forming a zoned icing warning, and issuing differentiated deicing control strategies to the deicing device LCU of each sub-area. For the shadow center, the deicing device starts quickly, adjusts the heating circuit current to the rated value (such as 1000A), and adjusts the spraying cycle of the anti-icing liquid spraying device to the shortest (such as 1min / time) and the spraying volume to the maximum (such as 1L / time). Through the combination of electric heating and chemical anti-icing, the conductor temperature is quickly raised from -5℃ to 0℃ within 30min. For the shadow transition zone, the heating circuit current is adjusted to 0.6 times the rated value, the anti-icing liquid spraying cycle is adjusted to 2min / time, and the spraying volume is adjusted to 0.5L / time, maintaining the conductor temperature at around -2℃. For the shadow edge, the heating device is not started temporarily, the anti-icing liquid spraying cycle is 5min / time, and the spraying volume is 0.2L / time, so that the conductor temperature is maintained at around -4℃. The deicing device of each sub-area corrects the control strategy in real time according to the movement of the cloud group. By actively inputting heat and chemicals, it accurately suppresses ice in the early stage of ice spread to avoid the aggravation of ice hazards.

[0083] The transmission line driven by the digital twin closed loop is continuously optimized and controlled. The operating status of the control equipment at each level is transmitted back in real time through the communication link. After being aggregated at the substation level, it is uploaded to the digital twin system through the backbone optical fiber or 5G network. In the voltage control scenario, the bus voltage of each substation, the apparent power of the line, the equipment action items and other data are presented in the form of voltage curves, power curves, alarm logs, etc. The digital twin system continuously calculates the line voltage qualification rate and evaluates the optimization control effect. When the voltage exceeds the limit again, it automatically corrects the optimization model and initiates a new round of multi-station and multi-device collaborative voltage control to achieve continuous improvement of the voltage qualification rate under the digital twin closed loop. In the icing control scenario, massive heterogeneous data such as micro-meteorological sensors, conductor temperature sensors, and deicing device energy consumption sensors along the line are transmitted back in real time through the low-power wide area network. The digital twin system analyzes the trend of ice temperature changes in each section and evaluates the differentiated deicing effect of the partition. Once it is found that the local ice temperature is still rising, the digital twin model automatically adjusts the deicing strategy of the local area to further increase the heating power and the amount of anti-icing fluid sprayed. At the same time, the de-icing section division is optimized in combination with the weather forecast to minimize the de-icing energy consumption cost. Through bottom-up data fusion and top-down strategy optimization, a closed-loop control of the ice-covered temperature of the transmission line is formed to ensure that the icing hazards are controlled throughout the process.

[0084] In summary, this paper proposes a digital twin-driven transmission line voltage and icing temperature coordinated optimization control strategy. Based on the abnormal warning of the digital twin system, this method formulates intelligent control schemes for voltage optimization and icing optimization for typical working conditions such as voltage over-limit and icing temperature rise. In voltage optimization control, hierarchical multi-station and multi-equipment coordination is adopted, and the voltage of key nodes of the line is accurately controlled to restore to a safe range through the optimal combination adjustment of transformers, reactors, and capacitors. In icing optimization control, the icing risk section is divided by shadow projection, and differentiated intelligent deicing is implemented according to the shadow position. By accurately adjusting the working parameters of the deicing device, the deicing energy consumption is minimized while suppressing the icing temperature rise. The voltage and icing optimization control process interacts with the digital twin model in real time. Through operating state mapping, control effect evaluation, and adaptive correction of strategy parameters, a full-cycle closed-loop control is constructed to form a new mode of safe and optimized operation of transmission lines continuously driven by digital twins. Subsequent research directions mainly include: further enriching the constraints of voltage optimization control, taking into account reactive power optimization configuration while ensuring the voltage qualification rate; combining energy feedback deicing technology to realize the reuse of deicing energy consumption in ice optimization control; comprehensively considering various disaster conditions such as strong wind vibration and lightning discharge, and establishing a coordinated prevention and control model for power transmission line disasters in complex environments. At the same time, it is also necessary to build a hardware-in-the-loop test platform to verify the feasibility and economy of the proposed collaborative optimization control strategy through physical simulation experiments, so as to help the engineering application of digital twins in the intelligent operation and maintenance of power transmission lines.

[0085] In an optional embodiment,

[0086] The optimal transmission line operation parameter combination obtained by optimizing the digital twin model is sent to the intelligent control system of the transmission line. The intelligent control system collaboratively controls the reactive compensation device, the dynamic voltage regulator, and the intelligent deicing device to dynamically adjust the reactive power, voltage amplitude and phase, and conductor temperature parameters of the transmission line so that the transmission line reaches the optimal operating state under the current environmental conditions and meets the power quality requirements. The steps include:

[0087] Based on the intelligent control system architecture, this paper further explores the application of multi-agent game theory in distributed collaborative optimization control of transmission lines. In view of the complexity of cross-regional multi-device collaborative control of large-scale transmission lines, a hierarchical multi-agent distributed optimization control framework is proposed. Configure regional control agents in each control partition, set up line control agents at the line level, and equip local control agents on intelligent control devices to form a three-layer agent structure of "device-region-line". Through the game theory mechanism modeling of strategic interaction and interest coordination among multiple agents, the agents converge to Nash equilibrium in autonomous learning and evolution, while pursuing the maximization of their own utility, taking into account the global power quality, and realizing distributed adaptive collaborative optimization control of transmission lines. The specific technical solution is as follows:

[0088] Construct a hierarchical distributed multi-agent control architecture. According to the geographical span and electrical topology of the transmission line, it is divided into several control zones, each of which is responsible for the autonomous control of the internal equipment. Taking a ±800kV UHV DC transmission project as an example, the transmission line is 2200km long and spans 5 provincial administrative regions. It can be divided into 12 control zones with a zone span of 150~200km. Configure an area control agent (ACAS) in each zone to parse the superior dispatching instructions and formulate a multi-device collaborative optimization plan for the region. Deploy the device control agent (DCAS) on local control devices such as intelligent switches, reactive power compensation devices, dynamic voltage regulators, and intelligent deicing devices to execute superior control instructions and drive primary equipment to carry out refined adaptive adjustment. Set up a line control agent (LCAS) at the line level to coordinate the cross-regional control behavior of ACAS. The three types of agents are logically associated according to the hierarchical distributed architecture, in which DCAS is affiliated to ACAS and accepts the dispatch of ACAS; multiple ACAS achieve cross-regional collaboration through LCAS, and LCAS is the superior dispatch of ACAS. All three types of intelligent agents adopt a multi-agent computing model, which can perceive the environment and make autonomous decisions based on it, and exchange status and strategy information with the digital twin system, sibling agents, and superior and subordinate agents through communication interfaces.

[0089] The regional control agent generates a partition multi-device optimization scheduling strategy. ACAS obtains the combination of target operating parameters of multiple devices in the region optimized by the digital twin model in real time, mainly including transformer tap position, capacitor and reactor switching status, conductor temperature, etc. At the same time, ACAS analyzes the real-time health level and available regulation capacity of the equipment in the region through state estimation to form a constraint set of available capacity of the equipment. Based on the multi-objective genetic algorithm, the weighted sum of line power quality, equipment health level, and regulation capacity margin is used as the fitness function, and the optimal scheduling combination of multiple devices is solved through mechanisms such as encoding, selection, crossover, and mutation. Taking a certain partition as an example, the genetic algorithm is used to obtain that the two transformers in the region are adjusted to the 8th and 11th gears respectively, two groups of capacitors are put into use, and the deicing device is started to maintain the conductor temperature at -2℃. Under the premise of meeting the reactive balance and ice prevention in the region, the comprehensive health index of the equipment can be increased by 12% compared with before optimization, and 8% of the voltage regulation margin is reserved to deal with future uncertainties. ACAS converts the optimization scheduling results of multiple devices into executable device-level control instructions, which are sent to the corresponding DCAS through the optical fiber communication link. DCAS analyzes the equipment control instructions and converts them into port drive signals, controls the action of the equipment actuator, and adjusts the equipment operation status to track the optimization target. The status and alarm information of each equipment during execution are fed back to ACAS in real time through DCAS, and ACAS corrects and optimizes the scheduling plan accordingly to achieve closed-loop autonomous control of multiple equipment in the area.

[0090] The line control agent builds a cross-regional coordination mechanism. The partitioned autonomous control of multiple ACAS may cause voltage imbalance and reactive power exchange between regions, which needs to be coordinated at the line level. LCAS keeps strategy synchronization with ACAS through an adaptive information interaction mechanism, obtains the partitioned multi-device optimization scheduling plan generated by ACAS, extracts key features such as bus voltage and line reactive power in each region, and evaluates the cross-regional power quality level. Once it is found that the voltage deviation between regions exceeds 1.5% or the reactive power exchange exceeds the threshold, LCAS switches to the cross-regional collaborative control mode. Taking the adjacent M and N areas as an example, the bus voltage of M area is 532kV and that of N area is 528kV. The voltage deviation between regions is 0.76%, the apparent power of the line is 600MVA, and the reactive power is 180MVar, which exceeds the cross-regional reactive exchange threshold of 150MVar. Based on the Powell iterative optimization algorithm, LCAS takes the minimization of voltage deviation and reactive power balance between the two areas as the optimization objectives, and quickly calculates the coordinated control correction of the two areas through iterative optimization, including a voltage reduction of 0.6kV in area M, a voltage increase of 0.5kV in area N, and a reactive power reduction of 30MVar from area M to area N. LCAS feeds the above corrections back to the ACAS in areas M and N as coordinated control instructions. While meeting the autonomous control of the two areas, the ACAS in the two areas further refines the superior dispatching instructions, formulates transformer taps and capacitor bank switching plans, and completes the regional control correction within the time limit specified by LCAS. Through the cross-area coordination of LCAS, the voltage deviation in the key control area of ​​the transmission line is reduced to 0.2%, and the reactive power of the line is reduced to 135MVar, meeting the qualified standards for cross-regional power quality. LCAS can also exchange key control information through the line control intelligent body of the friendly station to coordinate the reactive balance and voltage imbalance across the line.

[0091] The game theory mechanism is introduced to model the multi-agent distributed collaborative game optimization process. Multiple ACAS are regarded as rational game parties in the distributed collaborative optimization game model. By designing a utility function that comprehensively considers control performance and global interests, the game parties are guided to form a stable equilibrium strategy in autonomous evolution. Taking the ACAS of M and N areas participating in cross-regional collaborative control as an example, the utility function is defined as the weighted sum of "power quality performance of this area, comprehensive health level of equipment, and contribution to the promotion of power quality in neighboring areas", with weight coefficients of 0.5, 0.3, and 0.2 respectively. Through the internal multi-equipment collaboration, the power quality qualification rate of M area reached 98%, but due to the downward adjustment of voltage instructions, the health level of transformers and capacitor groups decreased by 2%. At the same time, in order to cooperate with the voltage recovery of N area, 30MVar of reactive power was actively transmitted to N area. After weighted calculation, the utility function output of M area is 0.93. Due to LCAS dispatching, the power quality level and equipment health level of N area are better than those of M area, but the contribution to the reactive support of M area is small, and the utility function output is 0.96. Through continuous strategic game, Zone M realized that it could appropriately increase the contribution weight to the power quality of Zone N, and actively increase the reactive power support for Zone N while moderately reducing the power quality requirements of users in this zone. After Zone N learned that Zone M had increased its support for it, it correspondingly increased the game return to Zone M and actively provided voltage support to Zone M. After multiple rounds of strategy adjustments, the utility function outputs of Zones M and N gradually converged and stabilized at around 0.95, reaching a game equilibrium. Through rational game, Zone M and Zone N ACAS formed an equilibrium strategy of "mutual benefit, win-win cooperation", which not only met the interests of the sub-areas, but also took into account the overall power quality, reflecting the self-organizing emergence characteristics of the intelligent body.

[0092] A distributed trust mechanism for multi-agent strategy interaction is constructed based on blockchain. In order to ensure the security and reliability of the cross-regional collaborative game process, the alliance chain technology is introduced to build a tamper-proof strategy interaction evidence chain. ACAS acts as a blockchain node. The information on the chain mainly includes the partition multi-device optimization scheduling plan submitted by ACAS, the cross-regional collaborative control instructions issued by LCAS, etc., which are encrypted and calculated through cryptographic mechanisms. ACAS organizes the strategy information to be synchronized into blocks, and records the hash value of the previous block in the block header to form a strategy interaction evidence chain. The tail block of the chain records the current latest strategy. ACAS competes for the right to record accounts through a consensus mechanism. The ACAS that obtains the right to record accounts is responsible for incorporating new blocks into the evidence chain. In order to control the number of accounting nodes and avoid excessive communication overhead caused by the participation of all ACAS in the consensus, a proof-of-stake consensus protocol such as DPoS or PBFT can be adopted. LCAS aggregates the strategy interaction blocks of ACAS and submits them to the accounting committee composed of partition representative nodes. After more than 2 / 3 of the members of the entire network vote in favor, the block confirmation is completed and broadcast to each ACAS. The distributed ledger based on blockchain can realize the functions of traceable storage, consistency verification, tampering detection of policy information, and build a trusted interactive environment for intelligent agents covering the entire line. If an attacker tampers with the local policy and creates inconsistencies with other node ledgers, it can be detected and identified by multi-RMI and then isolated. At the same time, LCAS, as a blockchain governance node, can trace improper gaming behavior and punish dishonest ACAS, motivating multiple intelligent agents to form a trustworthy and mutually beneficial strategic game mechanism, and realize the trusted security of distributed collaborative optimization control of transmission lines.

[0093] In summary, this paper proposes a distributed collaborative optimization control strategy for transmission lines based on multi-agent game. According to the geographical and electrical characteristics of the line, multiple control zones are divided to construct a three-layer agent hierarchical architecture of "equipment-region-line", in which the regional control agent is responsible for parsing the line-level dispatching instructions to form a partitioned multi-device optimization strategy, which is executed in a closed loop through the equipment control agent. The line control agent evaluates the global power quality level through cross-regional information interaction, and forms collaborative control instructions to guide regional control behavior. The game theory mechanism is introduced to model the distributed collaborative optimization game process of multi-regional control agents. By designing a comprehensive utility function of "self-performance and global contribution", the game parties are encouraged to form a mutually beneficial and win-win equilibrium strategy in autonomous evolution. Based on blockchain, a trusted environment for agent strategy interaction is constructed to ensure the integrity and credibility of multi-regional collaborative games, and a self-organizing, adaptive, and self-evolving distributed collaborative optimization control of transmission lines with emergent characteristics is realized. Subsequent research directions mainly include: further optimizing cross-regional coordination mechanisms and shortening the strategy convergence time of multi-agents; exploring the application of deep reinforcement learning in utility function design and improving the stability of equilibrium strategies; combining sociological mechanisms to design reasonable incentives and build a credible cross-regional benefit distribution and compensation mechanism. At the same time, it is also necessary to carry out hardware and software integration testing to evaluate the engineering applicability of the proposed distributed collaborative optimization control method and provide theoretical and technical support for the large-scale deployment of intelligent transmission lines.

[0094] Figure 2 Schematic diagram of the structure of the power quality monitoring and intelligent optimization system driven by the digital twin according to the embodiment of the present invention. Figure 2 As shown, the system comprises:

[0095] The first unit is used to construct a three-dimensional geometric model, material attribute model and physical property model of the transmission line, and establish a digital twin model of the transmission line by fusing multi-source heterogeneous data; wherein the three-dimensional geometric model includes the geometric parameters of the tower, conductor, insulator string and hardware transmission line, the material attribute model includes the material model of the conductor, the material parameters of the dielectric constant and dissipation factor of the insulating material, and the physical property model includes the electrical parameters of the resistivity, inductance and capacitance of the conductor, and the physical parameters of the mechanical strength of the tower and insulator string; a power quality monitoring unit is arranged at the measuring point of the transmission line, and the power quality monitoring unit includes a high-precision voltage transformer, a current transformer, a harmonic analyzer, and a temperature and humidity sensor. The power quality monitoring unit uploads the power quality data collected and analyzed to the edge computing server in real time through the optical fiber private network; the edge computing server pre-processes the power quality data, evaluates the health status of the power quality of the transmission line through an intelligent algorithm, generates a power quality evaluation report and uploads it to the cloud server;

[0096] The second unit is used for the cloud server to receive the power quality assessment report uploaded by the edge computing server. If the power quality assessment report shows that the power quality does not meet the standard, the digital twin model is called to simulate the process of power quality optimization of the transmission line in the virtual space by adjusting the parameters of the digital twin model. On the basis of solving the power flow equation group, a multi-objective optimization algorithm is used to search for the optimal combination of transmission line operation parameters that meet the power quality standard, including the conductor cross-sectional area, reactive power compensation capacity, and ambient temperature and humidity control threshold;

[0097] The third unit is used to send the optimal transmission line operating parameter combination obtained by optimizing the digital twin model to the intelligent control system of the transmission line. The intelligent control system collaboratively controls the reactive compensation device, the dynamic voltage regulator, and the intelligent de-icing device, and dynamically adjusts the reactive power, voltage amplitude and phase, and conductor temperature parameters of the transmission line to enable the transmission line to achieve the optimal operating state under the current environmental conditions and meet the power quality requirements. The intelligent control system feeds back the optimized control effect to the digital twin model for correcting and updating the relevant parameters of the digital twin model.

[0098] According to a third aspect of the embodiments of the present invention,

[0099] An electronic device is provided, comprising:

[0100] processor;

[0101] a memory for storing processor-executable instructions;

[0102] The processor is configured to call the instructions stored in the memory to execute the aforementioned method.

[0103] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. The power quality monitoring and intelligent optimization method based on digital twin is characterized by: include: Construct a three-dimensional geometric model, material property model and physical property model of the transmission line, and establish a digital twin model of the transmission line by fusing multi-source heterogeneous data; wherein the three-dimensional geometric model includes geometric parameters of the tower, conductor, insulator string and hardware transmission line, the material property model includes the material model of the conductor, the material parameters of the dielectric constant and dissipation factor of the insulating material, and the physical property model includes the electrical parameters of the resistivity, inductance and capacitance of the conductor, and the physical parameters of the mechanical strength of the tower and insulator string; arrange a power quality monitoring unit at the measuring point of the transmission line, the power quality monitoring unit includes a high-precision voltage transformer, a current transformer, a harmonic analyzer, and a temperature and humidity sensor, and the power quality monitoring unit uploads the collected and analyzed power quality data to the edge computing server in real time through the optical fiber private network; the edge computing server pre-processes the power quality data, evaluates the health status of the power quality of the transmission line through an intelligent algorithm, generates a power quality evaluation report and uploads it to the cloud server; The cloud server receives the power quality assessment report uploaded by the edge computing server. If the power quality assessment report shows that the power quality does not meet the standard, the digital twin model is called to simulate the power quality optimization process of the transmission line in the virtual space by adjusting the parameters of the digital twin model. On the basis of solving the power flow equation group, a multi-objective optimization algorithm is used to search for the optimal transmission line operation parameter combination that meets the power quality standard, including the conductor cross-sectional area, reactive power compensation capacity, and ambient temperature and humidity control threshold. The optimal transmission line operating parameter combination obtained by optimizing the digital twin model is sent to the intelligent control system of the transmission line. The intelligent control system collaboratively controls the reactive compensation device, the dynamic voltage regulator, and the intelligent de-icing device, and dynamically adjusts the reactive power, voltage amplitude and phase, and conductor temperature parameters of the transmission line, so that the transmission line reaches the optimal operating state under the current environmental conditions and meets the power quality requirements. The intelligent control system feeds back the optimized control effect to the digital twin model for correcting and updating the relevant parameters of the digital twin model.

2. The method according to claim 1, characterized in that The steps of building a 3D geometric model, material property model, and physical property model of the transmission line and fusing multi-source heterogeneous data to establish a digital twin model of the transmission line include: Collect 3D point cloud data of transmission line towers, conductors, insulator strings, and hardware, and perform denoising, registration, and fusion processing on the collected 3D point cloud data to improve the quality of 3D point cloud data; use a combination of parametric modeling and instance-based modeling to build a parametric 3D solid model, and generate 3D entities of different specifications through the instantiation method; use subdivision surface technology to perform personalized modeling for insulator strings and hardware; assemble the parametric 3D solid model into a complete 3D geometric model of the transmission line according to the spatial position relationship, and use scene tree and level of detail technology to optimize the organizational structure and rendering performance of the 3D geometric model; According to the metal material grade of the conductor, the physical property parameters of the conductor material are obtained from the material database, including density, resistivity, temperature coefficient, and stress-strain curve, and a multi-physics field constitutive model of the conductor material is established; by testing the dielectric constant, dissipation factor, and breakdown voltage of the insulator string dielectric material, an insulator string dielectric material model that reflects the frequency and temperature dependence of the insulator string dielectric material is constructed; for the materials of the tower, foundation, and hardware, the corresponding material model is selected from the material database according to their material composition and mechanical performance indicators to construct a material property model for the transmission line; On the basis of material property model and three-dimensional geometric model, the multi-physics coupling model of conductor is established by comprehensively considering the variation of resistivity, inductance and capacitance of conductor with temperature and stress; in view of surface flashover and pollution flashover of insulator string, the multi-physics model of insulator string is constructed to quantitatively simulate the dielectric properties and discharge characteristics of insulator string under different working conditions and environmental factors; according to the material properties and structural characteristics of the tower, the mechanical model of the tower is established, and the finite element method is used to analyze the strength, stiffness and stability of the tower under wind load and ice load conditions; the multi-physics coupling model of conductor, the multi-physics model of insulator string and the mechanical model of tower are integrated to construct the physical characteristic model of transmission line; The physical property model and material attribute model of the transmission line are integrated, a three-dimensional geometric model of the transmission line is constructed in a three-dimensional virtual environment, the physical property model, the material attribute model and the three-dimensional geometric model are matched through a mapping method, and a digital twin model of the transmission line is established.

3. The method according to claim 1, characterized in that The cloud server receives the power quality assessment report uploaded by the edge computing server. If the power quality assessment report shows that the power quality does not meet the standard, the digital twin model is called, and the process of optimizing the power quality of the transmission line is simulated in the virtual space by adjusting the parameters of the digital twin model. On the basis of solving the power flow equation group, the multi-objective optimization algorithm is used to search for the optimal combination of transmission line operating parameters that meet the power quality standard, including the following steps: Dynamically link the power quality assessment report received by the cloud server with the digital twin model, and accurately map the power quality issues reflected in the power quality assessment report to the corresponding parameters and state variables of the digital twin model by establishing a standardized data mapping mechanism; In response to the situation where the power quality exceeds the standard, a multi-scenario and multi-objective optimization control problem is constructed, which comprehensively considers the optimization objectives of power quality indicators, transmission line thermal capacity margin, and network loss, and combines load forecasting and new energy output forecasting data to generate constraints under multiple operating scenarios; The network topology, line parameters, and node type information of the transmission line in the digital twin model are converted into the standard format of the power flow equation group, and the power flow equation group is solved using the fast decoupling method to obtain the voltage, current, and power distribution of the transmission line under different operating parameters; Taking into account the conflict and coupling between power quality indicators, a multi-objective evolutionary algorithm is used to solve multi-scenario and multi-objective optimization problems. Problem knowledge is integrated into the algorithm to improve the selection, crossover, and mutation operations, and the Pareto optimal solution set is searched. According to the Pareto optimal solution set, the parameters of the digital twin model are tuned. Through parameter sensitivity analysis and simulation verification, the optimal parameter combination that can achieve power quality standards under multi-scenario constraints is screened out, including reactive compensation capacity, tap adjustment position, and new energy output limit; The optimal parameter combination output by the digital twin model is decoded into control instructions for the physical transmission line. The two-way mapping between the digital twin model and the physical transmission line is achieved through a standard interface. The physical transmission line is controlled in real time, and the power quality and equipment operating status are continuously monitored. The feedback information is synchronized to the digital twin model for closed-loop optimization.

4. The method according to claim 3, characterized in that In response to the situation where the power quality exceeds the standard, a multi-scenario and multi-objective optimization control problem is constructed, and the optimization objectives of power quality indicators, transmission line thermal capacity margin, and network loss are comprehensively considered. In addition, the load forecast and new energy output forecast data are combined to generate the constraints under multiple operation scenarios. The steps include: According to the power quality assessment report, a weighted multi-objective function is constructed with the sum of squares of voltage deviation, the sum of squares of harmonic distortion rate deviation, the sum of squares of line current margin, network loss and reactive power optimization as sub-objectives, and the priorities between the sub-objectives are balanced by setting weight coefficients. Establish multi-dimensional constraints that take into account the physical constraints and safety constraints of the transmission line, including equality constraints formed by node power balance equations, inequality constraints formed by bus voltage amplitude and line current constraints, and control variable constraints formed by power quality index limits and equipment physical adjustment ranges; An intelligent optimization algorithm is used to solve the multi-scenario and multi-objective transmission line optimization model. The power flow calculation program is called in each scenario, and the state variables are substituted into the objective function and constraints. The constraint processing technology of the penalty function method is used to solve the problem into an unconstrained multi-objective optimization problem, and the optimal control strategy that meets the power quality standards is searched.

5. The method according to claim 1, characterized in that The cloud server receives the power quality assessment report uploaded by the edge computing server. If the power quality assessment report shows that the power quality does not meet the standard, the digital twin model is called to simulate the process of power quality optimization of the transmission line in the virtual space by adjusting the parameters of the digital twin model. On the basis of solving the power flow equation group, a multi-objective optimization algorithm is used to search for the optimal transmission line operation parameter combination that meets the power quality standard, including the conductor cross-sectional area, reactive power compensation capacity, and ambient temperature and humidity control threshold. The steps include: A hybrid coding strategy is used to encode the control variables of the transmission line, and the initial particle group is randomly generated. The power flow calculation module is used to calculate multiple power quality indicators corresponding to each particle, and a dimensionless weighted fitness function is constructed as a particle evaluation index. The initial particle swarm is non-dominated sorted, particles are stratified according to Pareto level, the crowding distance of particles is calculated, the selection operator integrating problem-related knowledge is used to select parent particles from the current particle swarm, and the crossover operation is performed in combination with the power quality index weight information. At the same time, heuristic rules are introduced to repair infeasible solutions. According to the sensitivity information of the power quality of the transmission line to the control variables, the mutation probability and the step length of the mutation are adaptively adjusted, the parent particles are mutated to generate child particles, the parent particles and the child particles are merged to form a new particle group, and the merged particle group is non-dominated sorted; Based on the Pareto level and crowding distance information of particles, excellent particles are selected to form a new generation of particle swarms. Non-dominated particles are preferentially retained in the external archive set to participate in particle fitness evaluation, and an adaptive weight adjustment strategy is introduced to dynamically update the weights of power quality indicators. The improved adaptive penalty function method is used to deal with constraint violations in the new generation of particle swarms. The fitness of the new generation of particles is penalized according to the degree of satisfaction of the power quality indicators. The standard particle swarm algorithm is used to iteratively update the continuous control variables, and the quantum behavior particle swarm algorithm is introduced to iteratively update the discrete control variables. The above particle swarm iterative evolution steps are repeated until the maximum evolutionary generation or particle swarm convergence requirements are met, and the particles in the external archive set are output as the Pareto optimal solution set of the multi-scenario transmission line optimization control problem. Each solution in the Pareto optimal solution set corresponds to a transmission line optimization control strategy that balances multiple power quality indicators.

6. The method according to claim 1, characterized in that The optimal transmission line operation parameter combination obtained by optimizing the digital twin model is sent to the intelligent control system of the transmission line. The intelligent control system collaboratively controls the reactive compensation device, the dynamic voltage regulator, and the intelligent deicing device to dynamically adjust the reactive power, voltage amplitude and phase, and conductor temperature parameters of the transmission line so that the transmission line reaches the optimal operating state under the current environmental conditions and meets the power quality requirements. The steps include: The digital twin model outputs the optimal combination of transmission line operating parameters under the current environmental conditions, including reactive power, voltage amplitude and phase, and conductor temperature, and transmits it to the master station layer of the intelligent control system through the optimization control instruction issuing interface; after receiving the optimization control instruction, the master station layer decomposes the optimization control instruction into a control strategy for the substation layer, and sends it to the smart substation terminal and the local intelligent controller; the substation layer formulates a multi-device coordinated control plan based on the received control strategy, comprehensively considers the adjustment capability, adjustment speed, and reliability factors of each controllable device, and calculates the equipment combination control action with the greatest comprehensive benefit; The substation-level intelligent controller converts the equipment combination control action into executable instructions and sends them to the primary equipment respectively, including transformers, static VAR compensators, dynamic voltage restorers, and intelligent deicing devices; the primary equipment executes the substation control instructions in parallel, and coordinates the parameters of reactive power, voltage amplitude, and conductor temperature of the transmission line by adjusting the transformer tap position, switching the capacitor bank tap, changing the output voltage of the power electronic device, and optimizing the operation mode of the deicing device; Intelligent sensors at the equipment layer collect operating data on voltage, current, power, power quality, and equipment health status in real time, and upload them to the substation layer for aggregation through a multi-hop wireless communication network; the substation layer uploads the collected equipment operating status data to the digital twin model through the equipment operating status synchronization interface for real-time correction, and feeds back the power quality monitoring data to the digital twin model through the power quality data upload interface at the main station layer; The main station layer continuously monitors the power quality level of the transmission line. When the power quality index exceeds the limit, the digital twin model is triggered to optimize again, and the intelligent alarm is activated to prompt the line abnormality; the digital twin model periodically obtains load power forecasts, new energy output forecasts, and weather forecast data along the line, dynamically updates internal parameters, and continuously optimizes in combination with power quality feedback data, outputs a new optimal transmission line operating parameter combination, and realizes continuous optimal operation of the transmission line through closed-loop feedback control.

7. The method according to claim 6, characterized in that The method further comprises: When the digital twin model detects that the voltage of the transmission line deviates from the optimal value and exceeds the threshold, it uses the optimal power flow algorithm to calculate the distribution plan of the voltage regulation amount at each key node, and forms a control instruction to be sent to the master station layer; The master station layer forwards the voltage regulation control instructions to the intelligent terminals of the relevant substations and converter stations. The intelligent terminals determine the voltage regulation amount of each device based on the characteristics of the equipment that can participate in voltage regulation at the station and the digital twin model instructions, and issue them to the local intelligent controllers. The local intelligent controller assigns weights to each device execution action that can participate in regulation from the two dimensions of benefit and cost, and uses the difference between the benefit weight and the cost weight as the comprehensive benefit of a single regulation action; the local intelligent controller combines the feasible regulation actions of multiple devices, calculates the comprehensive benefit of each combined action, selects the combined action with the largest benefit as the optimal voltage regulation scheme, and sends the regulation instructions of each device in the combined action to the device layer; the actuators at the device layer complete their respective regulation actions in parallel, and collaboratively achieve the optimal regulation of the overall line voltage by changing the transformer tap position, switching reactors and capacitor bank taps; When the digital twin system predicts that the temperature rise of the ice-covered transmission line conductor exceeds the warning value, a de-icing control instruction is sent to the intelligent control system. The instruction includes the target de-icing section and de-icing temperature. The shadow position and boundary of the ice-covered section are tracked, and the de-icing section is divided into the shadow center, shadow transition zone and shadow edge, forming a partitioned de-icing task and sending it to the local controller of each de-icing device. The de-icing device located in the shadow center starts quickly, adjusts the heating circuit current to the rated value, and adjusts the anti-icing liquid spraying cycle and spraying amount to the maximum, so that the temperature of the ice-covered conductor rises rapidly to the target value. The de-icing device located in the shadow transition zone adjusts the heating current to 0.6 times the rated value, and the anti-icing liquid spraying cycle and spraying amount to medium, so as to maintain the conductor temperature slightly higher than the ambient temperature. The de-icing device located at the edge of the shadow does not start the heating circuit for the time being, but only maintains the conductor temperature slightly higher than the ambient temperature, and adjusts the anti-icing liquid spraying cycle and spraying amount to the minimum. The conductor temperature sensor uploads the real-time temperature data to the substation, and the substation evaluates the overall de-icing effect and feeds it back to the digital twin model.

8. The method according to claim 1, characterized in that The optimal transmission line operation parameter combination obtained by optimizing the digital twin model is sent to the intelligent control system of the transmission line. The intelligent control system collaboratively controls the reactive compensation device, the dynamic voltage regulator, and the intelligent deicing device to dynamically adjust the reactive power, voltage amplitude and phase, and conductor temperature parameters of the transmission line so that the transmission line reaches the optimal operating state under the current environmental conditions and meets the power quality requirements. The steps include: Divide the transmission line into multiple control zones according to its geographical area and electrical topology. Configure a regional control agent in each zone, set up a line control agent at the transmission line layer, and equip the intelligent control device with a local control agent, forming a "device-region-line" three-layer agent hierarchical architecture. The regional control agent obtains the optimal transmission line operation parameter combination of multiple devices in the region output by the digital twin model, combines the real-time status of the equipment and environmental conditions, and uses a multi-objective intelligent optimization algorithm to generate an equipment-level optimization scheduling strategy that takes into account the health level and regulation capability of the equipment. The scheduling command is sent to the local control agent and its feedback information is received to achieve closed-loop control. The line control agent coordinates the partition control behaviors of multiple regional control agents through an adaptive information interaction mechanism, summarizes the power quality assessment reports of each region and solves the cross-regional multi-objective collaborative optimization model, feeds back the regional control correction amount as a collaborative control instruction to the regional control agent, and realizes cross-regional reactive power balance and voltage coordination through cross-line information interaction; The idea of ​​game theory is introduced to construct a distributed collaborative optimization game model based on multi-agents. The regional control agent is used as a rational player and the utility function and incentive mechanism are designed. Through repeated games, multi-agents can take into account the overall interests while pursuing their own utility maximization, guide the autonomous evolution of multi-agent strategies to converge to Nash equilibrium, and improve the adaptability and robustness of collaborative control. A distributed trust mechanism based on blockchain technology is used to build a tamper-proof strategy interaction evidence chain for the multi-agent game process, to ensure the security and reliability of information synchronization between agents and the game process, and to achieve reliable distributed multi-agent line collaborative optimization control.

9. A digital twin-driven power quality monitoring and intelligent optimization system, used to implement the method described in any one of claims 1 to 8, characterized in that: include: The first unit is used to construct a three-dimensional geometric model, material attribute model and physical property model of the transmission line, and establish a digital twin model of the transmission line by fusing multi-source heterogeneous data; wherein the three-dimensional geometric model includes the geometric parameters of the tower, conductor, insulator string and hardware transmission line, the material attribute model includes the material model of the conductor, the material parameters of the dielectric constant and dissipation factor of the insulating material, and the physical property model includes the electrical parameters of the resistivity, inductance and capacitance of the conductor, and the physical parameters of the mechanical strength of the tower and insulator string; a power quality monitoring unit is arranged at the measuring point of the transmission line, and the power quality monitoring unit includes a high-precision voltage transformer, a current transformer, a harmonic analyzer, and a temperature and humidity sensor. The power quality monitoring unit uploads the power quality data collected and analyzed to the edge computing server in real time through the optical fiber private network; the edge computing server pre-processes the power quality data, evaluates the health status of the power quality of the transmission line through an intelligent algorithm, generates a power quality evaluation report and uploads it to the cloud server; The second unit is used for the cloud server to receive the power quality assessment report uploaded by the edge computing server. If the power quality assessment report shows that the power quality does not meet the standard, the digital twin model is called to simulate the process of power quality optimization of the transmission line in the virtual space by adjusting the parameters of the digital twin model. On the basis of solving the power flow equation group, a multi-objective optimization algorithm is used to search for the optimal combination of transmission line operation parameters that meet the power quality standard, including the conductor cross-sectional area, reactive power compensation capacity, and ambient temperature and humidity control threshold; The third unit is used to send the optimal transmission line operating parameter combination obtained by optimizing the digital twin model to the intelligent control system of the transmission line. The intelligent control system collaboratively controls the reactive compensation device, the dynamic voltage regulator, and the intelligent de-icing device, and dynamically adjusts the reactive power, voltage amplitude and phase, and conductor temperature parameters of the transmission line to enable the transmission line to achieve the optimal operating state under the current environmental conditions and meet the power quality requirements. The intelligent control system feeds back the optimized control effect to the digital twin model for correcting and updating the relevant parameters of the digital twin model.

10. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor is configured to call the instructions stored in the memory to execute the method described in any one of claims 1 to 8.

Images