A power system voltage stability collaborative optimization control method

Abstract

The application relates to the technical field of smart grid monitoring, and discloses a power system voltage stability cooperative optimization control method, which comprises the following steps: collecting transient evolution rates of controlled nodes and extracting monitoring channel pressure indexes, establishing a coupling criterion of the two to determine instruction dimensionality reduction reconstruction conditions; when the conditions are met, reconstructing full-precision numerical instructions into compressed instruction vectors containing adjustment direction bits and sensitivity level bits, and preemptively transmitting the compressed instruction vectors based on transmission priorities; and responding terminals restoring transient voltage compensation increments based on local cache data, and then driving a reactive power compensation unit. The application breaks the dependence of control precision on communication bandwidth, eliminates the mismatch between information time delay and physical evolution window, realizes the cooperative alignment of instruction transmission time scales and system evolution time sequences, suppresses adjustment oscillation induced by signal lag, and guarantees voltage stability under extreme communication environment.

Description

Technical Field

[0001] This invention belongs to the field of smart grid monitoring technology, and in particular relates to a collaborative optimization control method for voltage stability of power systems. Background Technology

[0002] Current power system voltage stability monitoring and control relies on real-time acquisition of node status parameters and the issuance of adjustment commands accordingly. With the increasing proportion of distributed power sources and nonlinear loads connected, the system operating status exhibits high-frequency fluctuations.

[0003] However, there is an inherent constraint between the physical evolution of the power system and the timing of information transmission. Increased monitoring frequency leads to increased load pressure on the communication link. Under limited channel bandwidth, control messages experience random queuing delays. This delay causes a deviation between the system state when the regulation command arrives at the execution terminal and the preset state when the command is generated, inducing regulation overshoot or control loop instability. Conventional methods of increasing bandwidth or computing power have not eliminated the mismatch between control loop delay and physical evolution window. Under extreme network congestion, there is still a risk of control failure. For example, Chinese invention patent CN119853079B discloses a method and device for main and distribution coordinated voltage optimization control under communication and electrical coupling. It uses multi-agent reinforcement learning to quickly solve the distributed voltage control model. This technology has advantages in handling complex nonlinear coupling. However, under extreme communication channel congestion or bandwidth-limited conditions, it relies heavily on high-frequency, full-precision state observation data to support decision-making logic. When message delays occur at the information layer, the model falls into closed-loop regulation oscillations due to the lack of timeliness of observation vectors. Furthermore, it lacks a differentiated transmission mechanism based on the urgency of the underlying physical evolution, making it difficult to achieve deterministic delivery of key control commands under extremely narrow bandwidth.

[0004] Therefore, the technical problem to be solved by this invention is how to construct a coordinated control mechanism with time consistency based on the coupling relationship between the communication channel state and the system physical evolution rate, so as to ensure voltage stability regulation under non-ideal communication environment. Summary of the Invention

[0005] This invention aims to solve the problem of voltage regulation failure caused by reduced timeliness of control messages under congested communication channel conditions.

[0006] In this technical solution, a collaborative optimization control method for voltage stability of a power system includes the following steps: Step S1: Collect the bus voltage sampling sequence of the controlled node, calculate the differential slope of the voltage sampling values ​​at adjacent sampling times, and obtain the transient evolution rate that characterizes the urgency of the state transition of the controlled node. Step S2: Extract the packet loss rate, round-trip time, and queue delay from the monitoring channel of the smart grid monitoring system, and calculate the monitoring channel pressure index that characterizes the degree of congestion in the monitoring channel. Step S3: Establish a coupling criterion between transient evolution rate and monitoring channel pressure index. By comparing the monitoring channel pressure index with a preset pressure threshold and the transient evolution rate with a preset rate threshold, determine whether the current control loop meets the instruction dimensionality reduction and reconstruction conditions. Step S4: If it is determined that the instruction dimensionality reduction and reconstruction conditions are not met, then a full-precision numerical control instruction is sent to the execution terminal to which the controlled node belongs. Step S5: If the condition for instruction dimensionality reduction and reconstruction is met, the adjustment direction is determined based on the full-precision numerical control instruction, the physical validity window corresponding to the current control cycle is calculated, the full-precision numerical control instruction is reconstructed into a compressed control instruction vector containing the adjustment direction sign bit and the sensitivity level bit, and the compressed control instruction vector is transmitted based on a preemptive tag. The compressed control instruction vector reaches the execution terminal within the physical validity window. Step S6: The execution terminal receives the compression control command vector, retrieves the voltage sampling value of the previous moment cached in the local register, restores the transient voltage compensation increment by combining it with the preset local sensitivity constant, and superimposes the transient voltage compensation increment into the locally stored action quantity history to drive the reactive power compensation unit connected to the execution terminal to change the output power.

[0007] Preferably, the process of calculating the monitoring channel pressure index in step S2 includes: step S21, constructing a weighted evaluation model that includes packet loss rate variables, round-trip time variables, and delay jitter variance variables; step S22, inputting the real-time collected communication performance parameters into the weighted evaluation model to obtain the monitoring channel pressure index; wherein, if the monitoring channel pressure index continuously exceeds the preset pressure threshold for 3 sampling periods, a channel congestion trigger signal is generated.

[0008] Preferably, the process of obtaining the transient evolution rate in step S1 includes: step S11, obtaining the absolute value of voltage deviation for 5 consecutive sampling points within a preset sliding time window; step S12, calculating the first time derivative of the absolute value of voltage deviation, and generating the transient evolution rate by combining it with a preset time decay factor.

[0009] Preferably, the coupling criteria in step S3 include: establishing a binary decision matrix consisting of a channel pressure dimension and a physical evolution dimension; when the monitoring channel pressure index indicates that the available bandwidth of the monitoring channel is less than 128kbps and the transient evolution rate exceeds the preset rate threshold, the collaborative optimization control method switches the control mode from the central station centralized calculation mode to the execution terminal local logic recovery mode.

[0010] Preferably, the method for generating the compression control command vector includes: mapping the transient evolution rate to a preset quantization range to determine the corresponding sensitivity level; the sensitivity level bit adopts a binary encoding method to instruct the execution terminal to adjust the step amplitude of the reactive power output within the physical validity window.

[0011] Preferably, in step S5, the preemptive transmission tag is used to guide the smart grid's communication gateway to identify the specific protocol header of the compressed control command vector. The communication gateway allocates a reserved transmission time slot for the compressed control command vector, and the compressed control command vector arrives at the execution terminal within a 20ms physical response window.

[0012] Preferably, the method further includes step S7, whereby after the monitoring channel pressure index falls below the preset pressure threshold, the control center issues a full-precision numerical control command to the execution terminal to perform closed-loop correction on the cumulative adjustment deviation generated by the execution terminal in the local logic recovery mode.

[0013] Preferably, the execution terminal is a remote terminal unit or a smart inverter distributed in the distribution network; the local register adopts a ring queue structure to dynamically store the voltage measurement values ​​and compensation action history of the most recent 5 sampling cycles.

[0014] Preferably, the collaborative optimization control method is applied to smart grids that integrate distributed power sources and nonlinear loads. By coupling the congestion status of the monitoring channel to the voltage control loop, the collaborative alignment of the control command transmission timescale and the power system physical evolution timescale is achieved.

[0015] Compared with existing technologies, the power system voltage stability collaborative optimization control method of the present invention has the following advantages: 1. In the collaborative optimization control of voltage stability in power systems, the dynamic trigger threshold model identifies redundant sampling processes in the monitoring system based on the coupling relationship between voltage deviation change and time decay term. This model suppresses false triggers caused by high-frequency oscillations, reduces the total amount of message transmission under unsteady conditions, alleviates queuing pressure on narrowband communication links, and ensures that critical control messages obtain sufficient bandwidth resources during sudden fluctuations.

[0016] 2. By monitoring the packet loss rate and message latency of the communication channel in real time, and in conjunction with the physical evolution rate characterized by the primary side voltage deviation change rate, a time-scale alignment criterion for the information layer delay and the physical layer evolution window is established. When the communication entropy exceeds the preset threshold and the physical evolution is in the fast window, the system automatically starts the instruction reconstruction mechanism to compress the high-precision numerical message into a feature vector composed of direction sign bits and sensitivity level bits, using a small information cost to obtain deterministic arrival of control actions within the physical effective time.

[0017] 3. After receiving the feature vector, the execution terminal uses the sampled data from the previous moment cached in the local register and the preset sensitivity constant to reconstruct the transient compensation amount with physical meaning through linear extrapolation. This process avoids the parameter verification link that depends on high bandwidth, eliminates the parsing delay after the instruction arrives, and keeps the reactive power compensation action on the physical side in phase with the transient demand of the bus voltage, effectively suppressing the closed-loop regulation oscillation induced by the lag of the control signal. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating the method for coordinated control of voltage stability in a power system according to the present invention. Figure 2 This is a logic diagram of instruction reconstruction and interactive feedback between the monitoring center and the execution terminal of this invention. Detailed Implementation

[0019] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0020] It should be noted that all directional and positional terms used in this invention, such as: up, down, left, right, front, back, vertical, horizontal, inner, outer, top, bottom, transverse, longitudinal, center, etc., are only used to explain the relative positional relationship and connection between components in a specific state (as shown in the accompanying drawings). They are only for the convenience of describing this invention and do not require that this invention be constructed and operated in a specific orientation. Therefore, they should not be construed as limiting this invention. In addition, the descriptions of "first," "second," etc., in this invention are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated.

[0021] In the description of this invention, unless otherwise explicitly specified and limited, the terms installation, connection, and linking should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections; they can refer to direct connections or indirect connections through an intermediate medium; they can refer to the internal connection of two components. For those skilled in the art, the specific meaning of the above terms in this invention can be understood according to the specific circumstances.

[0022] In the description of this specification, references to the terms "an embodiment," "some embodiments," "illustrative embodiments," "examples," "specific examples," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example, and the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0023] A collaborative optimization control method for voltage stability in a power system includes the following steps: Step S1: Collect the bus voltage sampling sequence of the controlled node, calculate the differential slope of the voltage sampling values ​​at adjacent sampling times, and obtain the transient evolution rate that characterizes the urgency of the state transition of the controlled node. Step S2: Extract the packet loss rate, round-trip time, and queue delay from the monitoring channel of the smart grid monitoring system, and calculate the monitoring channel pressure index that characterizes the degree of congestion in the monitoring channel. Step S3: Establish a coupling criterion between transient evolution rate and monitoring channel pressure index. By comparing the monitoring channel pressure index with a preset pressure threshold and the transient evolution rate with a preset rate threshold, determine whether the current control loop meets the instruction dimensionality reduction and reconstruction conditions. Step S4: If it is determined that the instruction dimensionality reduction and reconstruction conditions are not met, then a full-precision numerical control instruction is sent to the execution terminal to which the controlled node belongs. Step S5: If the condition for instruction dimensionality reduction and reconstruction is met, the adjustment direction is determined based on the full-precision numerical control instruction, the physical validity window corresponding to the current control cycle is calculated, the full-precision numerical control instruction is reconstructed into a compressed control instruction vector containing the adjustment direction sign bit and the sensitivity level bit, and the compressed control instruction vector is transmitted based on a preemptive tag. The compressed control instruction vector reaches the execution terminal within the physical validity window. Step S6: The execution terminal receives the compression control command vector, retrieves the voltage sampling value of the previous moment cached in the local register, restores the transient voltage compensation increment by combining it with the preset local sensitivity constant, and superimposes the transient voltage compensation increment into the locally stored action quantity history to drive the reactive power compensation unit connected to the execution terminal to change the output power.

[0024] Preferably, the process of calculating the monitoring channel pressure index in step S2 includes: step S21, constructing a weighted evaluation model that includes packet loss rate variables, round-trip time variables, and delay jitter variance variables; step S22, inputting the real-time collected communication performance parameters into the weighted evaluation model to obtain the monitoring channel pressure index; wherein, if the monitoring channel pressure index continuously exceeds the preset pressure threshold for 3 sampling periods, a channel congestion trigger signal is generated.

[0025] Preferably, the process of obtaining the transient evolution rate in step S1 includes: step S11, obtaining the absolute value of voltage deviation for 5 consecutive sampling points within a preset sliding time window; step S12, calculating the first time derivative of the absolute value of voltage deviation, and generating the transient evolution rate by combining it with a preset time decay factor.

[0026] Preferably, the coupling criteria in step S3 include: establishing a binary decision matrix consisting of a channel pressure dimension and a physical evolution dimension; when the monitoring channel pressure index indicates that the available bandwidth of the monitoring channel is less than 128kbps and the transient evolution rate exceeds the preset rate threshold, the collaborative optimization control method switches the control mode from the central station centralized calculation mode to the execution terminal local logic recovery mode.

[0027] Preferably, the method for generating the compression control command vector includes: mapping the transient evolution rate to a preset quantization range to determine the corresponding sensitivity level; the sensitivity level bit adopts a binary encoding method to instruct the execution terminal to adjust the step amplitude of the reactive power output within the physical validity window.

[0028] Preferably, in step S5, the preemptive transmission tag is used to guide the smart grid's communication gateway to identify the specific protocol header of the compressed control command vector. The communication gateway allocates a reserved transmission time slot for the compressed control command vector, and the compressed control command vector arrives at the execution terminal within a 20ms physical response window.

[0029] Preferably, the method further includes step S7, whereby after the monitoring channel pressure index falls below the preset pressure threshold, the control center issues a full-precision numerical control command to the execution terminal to perform closed-loop correction on the cumulative adjustment deviation generated by the execution terminal in the local logic recovery mode.

[0030] Preferably, the execution terminal is a remote terminal unit or a smart inverter distributed in the distribution network; the local register adopts a ring queue structure to dynamically store the voltage measurement values ​​and compensation action history of the most recent 5 sampling cycles.

[0031] Preferably, the collaborative optimization control method is applied to smart grids that integrate distributed power sources and nonlinear loads. By coupling the congestion status of the monitoring channel to the voltage control loop, the collaborative alignment of the control command transmission timescale and the power system physical evolution timescale is achieved.

[0032] Example 1: In an active distribution network operation scenario involving a high proportion of distributed power sources and high-power charging pile loads, when the system faces voltage fluctuations caused by the random access of large-scale nonlinear loads, the method of this invention maintains power system voltage stability in the following way: The smart grid monitoring system continuously collects the bus voltage sampling sequence of the controlled nodes, calculates the differential slope of the voltage sampling values ​​at adjacent sampling times, and obtains the transient evolution rate characterizing the urgency of the state transition of the controlled nodes. The monitoring system utilizes monitoring logic based on event trigger sensitivity coordination to construct a dynamic trigger threshold model jointly defined by the basic deviation and the time decay term. The dynamic trigger threshold model is determined by the event trigger function. Make a judgment: ,in, The value of the event triggering function; This represents the absolute value of the voltage deviation at the current moment. The baseline deviation threshold is set to 0.05. This is the dynamic attenuation coefficient, with a value of 0.1. It is the base of the natural logarithm; This is the time decay constant, and its value is 5; This is the current sampling time; The time when the last event was triggered.

[0033] when When the value is greater than 0, a trigger event is generated, and the system switches from normal monitoring to collaborative response mode. The monitoring system extracts the packet loss rate, round-trip time, and queue delay from the monitoring channel, calculates the monitoring channel pressure index that characterizes the degree of congestion in the monitoring channel, and simultaneously retrieves the local voltage and reactive power sensitivity matrix preset in the local memory. In actual calculations, the local voltage and reactive power sensitivity matrix... It is a square matrix constructed based on the partial derivatives of the voltage of all compensation nodes in the controlled area with respect to the change in reactive power. This is the controlled node voltage deviation vector obtained at the current sampling time. The execution terminal calculates the inverse matrix of the local voltage and the reactive power sensitivity matrix. With voltage deviation vector The product of these terms yields the transient voltage compensation increment vector corresponding to each compensation unit. Each component in this vector corresponds to the reactive power adjustment step size of a specific compensation unit, thus ensuring the dimensional consistency of instructions and the accuracy of algebraic operations in a multi-node collaborative environment. To balance the monitoring accuracy requirements with channel throughput, the system establishes a coupled criterion between transient evolution rate and monitoring channel pressure index, and uses the monitoring channel pressure index to extract the average dwell time of control messages to be sent in the queue. The physical effectiveness window of the current regulation command is calculated based on the rate of change of voltage deviation at the controlled node. ,when and The proportional relationship satisfies the condition for instruction dimensionality reduction and reconstruction, that is... and When the ratio exceeds 0.8, the control center reconstructs the full-precision numerical control command into a compressed control command vector containing the adjustment direction sign bit and the sensitivity level bit, and transmits the compressed control command vector based on a preemptive transfer tag.

[0034] Before inputting the collected communication performance parameters into the weighted evaluation model, the communication gateway processor of the smart grid monitoring system performs dimensionless operations on each parameter according to the range transformation algorithm, extracts the real-time measured values ​​of packet loss rate, round-trip delay, and delay jitter variance, subtracts the corresponding channel historical measurement minimum value pre-stored in the local database to obtain the first difference value, divides it by the physical interval width formed by the historical maximum and minimum values ​​to output the corresponding normalized variable, and multiplies each normalized variable by the set weight coefficient to generate the monitoring channel pressure index. Based on the preemptive transmission tag underlying physical processing logic, when the communication gateway encapsulates a transmission message containing a compressed control command vector, it directly sets the priority code point field of the VLAN tag in the Ethernet frame header to the highest level, level seven, or rewrites the differential service code point in the network layer header to an accelerated forwarding classification. This enables the media access control layer of the network switching equipment along the way to trigger the internal hardware queue priority scheduling action by parsing specific protocol fields. Under this scheduling action, the network switching equipment temporarily suspends the forwarding queue of the current regular business background data stream to ensure the time determinism of the compressed control command vector reaching the execution terminal.

[0035] The compression control command vector arrives at the execution terminal within a 20ms physical response window. The execution terminal receives and parses the compression control command vector, and uses the existing remote terminal unit in the controlled node to execute the control program. The execution terminal retrieves the voltage sampling value from the previous moment cached in the local register, and combines it with a preset local sensitivity constant to reconstruct the transient voltage compensation increment. : ,in, The recovered transient voltage is compensated for by an increment; The total number of reactive power compensation units participating in the adjustment; For cell index; According to the first The response weight is determined by the remaining capacity of each reactive power compensation unit; For the first The inverse matrix of the reactive power-voltage sensitivity matrix corresponding to each unit; For the voltage deviation monitored at the current moment, the execution terminal adds the restored transient voltage compensation increment to the locally stored action history, directly driving the connected reactive power compensation unit to change its output power. This process reduces the consumption of communication bandwidth through instruction dimensionality reduction and eliminates the queuing delay of instructions when the channel is congested. When the monitored channel pressure index falls below the preset pressure threshold, the control center issues a full-precision numerical control instruction to perform closed-loop correction on the cumulative adjustment deviation generated by the execution terminal in the local logic recovery mode. The actual bus voltage sampling trajectory continuously recorded and uploaded to the system historical database during the execution terminal's local logic recovery mode is extracted, and the discrete-time integral error between the trajectory and the steady-state reference voltage point is calculated. This error is then combined with the local voltage and reactive power sensitivity matrix to calculate... The steady-state error compensation term is derived; after the terminal receives the full-precision numerical control command containing the vector superposition value, it is parsed through the internal communication interface and directly written to the reference register of the underlying power control inverter of the reactive power compensation unit. At the hardware level, the compensation action quantity based on the transient evolution rate extrapolation and restoration retained in the local ring queue structure is forcibly overridden. A full-precision writing process for the hardware-level register based on the discrete-time integral error is set to smooth out the voltage static deviation accumulated by the local execution of the dimension reduction command during communication congestion. After triggering the coordinated control, the bus voltage fluctuation returns to the safe range of 0.95pu to 1.05pu. The message transmission volume of the monitoring system is reduced by more than 65% compared with the fixed period sampling mode, thereby maintaining the voltage stability of the power system in the environment of limited communication resources.

[0036] Example 2: In a physical simulation experimental platform equipped with a 10kV active distribution link, including a 3.5MW distributed photovoltaic array and 50 high-power DC charging piles, the closed-loop interaction of the smart grid monitoring system was simulated through a distributed computing unit. The experimental data came from the synchronous phasor measurement unit of the platform. The voltage sampling accuracy of the measurement unit was better than 0.1%, and the original data sampling frequency was set to 100Hz. The setting of the sampling period needs to balance the granularity of voltage transient capture and the channel bandwidth load. If the sampling frequency is lower than 50Hz, the voltage drop front when the nonlinear load is connected cannot be reproduced, while higher than 200Hz will lead to an increase in the monitoring message congestion rate. In this experiment, 100Hz was selected as the sampling reference. In order to simulate the industrial electromagnetic environment, Gaussian white noise with a signal-to-noise ratio of 35dB was superimposed on the sampling signal, and the 5th and 7th power frequency interference harmonics generated by the nonlinear load were simulated.

[0037] The experiment simulated a large-scale load surge by setting up a load spike scenario. Control group A used fixed-period sampling and full-precision command issuance. With the monitoring channel bandwidth limited to 512kbps, the monitoring channel pressure index remained above 0.85 due to the large payload of the full-precision messages and the fixed transmission frequency. This resulted in an average delay of over 120ms from control command generation to delivery to the terminal, and the bus voltage of the controlled node dropped to a minimum of 0.88pu. Secondary regulation oscillations occurred because the adjustment commands lagged behind the physical state. Control group B, while maintaining other consistent characteristics, removed the command dimensionality reduction and reconstruction step, implementing only event-triggered message reduction. Observational data showed that during the channel pressure surge phase, because the size of a single control message remained at 128 bytes, the average latency of the pending commands in the queue was [not specified]. The time reached 48.2 ms, exceeding the physical effectiveness window determined by the physical evolution rate. The time for the voltage fluctuation to return to the safe range is 2.6 seconds.

[0038] This invention applies a complete collaborative optimization control method for power system voltage stability. When a sudden change is detected in the bus voltage sampling sequence of the controlled node, the system activates a dynamic trigger threshold model, and the event trigger function... Under the judgment, the system only Exceeding the baseline deviation threshold When a threshold is reached, together with the dynamic attenuation factor, a monitoring message is sent. The monitoring channel pressure index extracted by the test group shows that, under bandwidth-constrained conditions, the control center uses this index to assess the risk of command failure. Specifically, when... and When the ratio reaches 0.82, the condition for instruction dimensionality reduction and reconstruction is met. The control center converts the full-precision numerical control instructions into a compressed control instruction vector of 16 bytes in length and assigns a preemptive transmission tag. Data records show that the transmission delay of the compressed instruction vector in the channel is reduced to 7.5ms. The execution terminal retrieves the voltage sample value from the local register cache and restores the transient voltage compensation increment by combining it with the preset local sensitivity constant. The reactive power compensation unit was driven to perform power compensation, limiting the bus voltage drop depth to 0.94 pu and shortening the voltage recovery time to 0.75 s. No power overshoot induced by regulation lag was observed. This was achieved for key parameters. Gradient verification shows that the technical effect of the present invention is logically related to the degree of deviation from the operating conditions. When the value is 0.05, the system maintains a trigger accuracy of 98.5% under background noise interference; when When reduced to 0.01, which is below the limit range and in the out-of-range control state, Gaussian white noise caused frequent false triggers, increasing the message throughput of the monitoring channel by 240%. The channel pressure index quickly saturated and caused the instruction queuing delay to run out of control. The test with increased problem intensity gradient showed that as the voltage fluctuation frequency increased from 1Hz to 10Hz, the bandwidth saving rate of the sample group of this invention increased from 45% to 72%, demonstrating the adaptive adjustment capability under complex fluctuation conditions. Through the synergy of instruction dimensionality reduction reconstruction and local data restoration, the mismatch between information delay and physical evolution time scale was eliminated, enabling the power system to maintain the bus voltage in the steady-state range of 0.95pu to 1.05pu under the environment of limited communication resources.

[0039] Example 3: This example combines Figures 1 to 2 This section describes a collaborative optimization control method for voltage stability in a power system, such as... Figure 1 As shown, step S1 involves acquiring the voltage sequence, calculating the differential slope to obtain the transient evolution rate, and then pointing downwards to step S2, which involves extracting channel features and calculating the channel pressure index representing congestion. Step S2 then points downwards to step S3, which involves determining the dimensionality reduction and reconstruction conditions based on the comparison between the rate and the pressure index. Step S3 then points downwards to step S4, which involves issuing a full-precision numerical control command when the reconstruction conditions are not met. Step S4 continues to point downwards to step S5, which involves reconstructing the compressed control command vector and preempting transmission when the conditions are met. Finally, step S5 points downwards to step S6 in the end-stage process, which involves restoring the compensation incremental drive reactive power unit to change its output power.

[0040] like Figure 2As shown in the diagram, the control center on the left is connected via solid lines (without arrows) to the functions of collecting the bus voltage sampling sequence of the controlled nodes, extracting monitoring channel pressure indicators, determining whether the current control loop meets the command dimensionality reduction and reconstruction conditions, and issuing full-precision numerical control commands. The communication gateway at the top of the diagram is connected via solid lines (without arrows) to the allocated reserved transmission time slots. In the control data flow relationship of the core components, both collecting the bus voltage sampling sequence of the controlled nodes and extracting monitoring channel pressure indicators are indicated by dashed arrows pointing to the determination of whether the current control loop meets the command dimensionality reduction and reconstruction conditions. Simultaneously, issuing full-precision numerical control commands is indicated by dashed arrows pointing upwards in reverse to this determination of whether the current control loop meets the command dimensionality reduction and reconstruction conditions. The fixed link is reconstructed into a compressed control command vector by a unidirectional dashed arrow, and the allocated reserved transmission time slots are also reconstructed into a compressed control command vector by dashed arrows. The reconstructed compressed control command vector leads to two output paths with dashed arrows. One dashed arrow points to the lower right to restore the transient voltage compensation increment, and the other dashed arrow turns back to the upper left to form feedback and points to the starting point to collect the bus voltage sampling sequence of the controlled node. In addition, the execution terminal is connected to the restored transient voltage compensation increment by a solid line without arrows. The restored transient voltage compensation increment then points downward by a solid arrow to change the output power. The reactive power compensation unit in the lower right of the figure is connected to the output power change by a solid line without arrows.

[0041] Example 4: In an active distribution network project with a large number of distributed power sources and flexible loads, when the system faces the coupled conditions of narrowing monitoring channel bandwidth and high-frequency voltage fluctuations caused by extreme weather, the method of this invention maintains the voltage stability of the power system through the following parameter calibration and protocol mapping. The smart grid monitoring system performs a weighted coefficient calibration procedure for the weighted evaluation model of the monitoring channel pressure index. Based on the electrical distance and importance of the controlled nodes in the distribution network, the distribution characteristics of the monitoring channel's contribution to voltage stability are extracted using offline power flow calculation. The weighted coefficients of the packet loss rate variable, the round-trip time variable, and the delay jitter variance variable are set to 0.4 and 0.2, respectively. If the monitoring channel pressure index obtained after processing the real-time collected communication performance parameters by the weighted evaluation model continuously exceeds the preset pressure threshold of 0.75 and reaches 3 sampling periods, the command reconstruction procedure of the control loop is triggered.

[0042] The system performs dynamic calibration based on physical evolution timescales for the proportional threshold in the instruction dimensionality reduction and reconstruction conditions, and uses the controlled nodes to calculate the minimum time constant for the controlled node bus voltage to migrate from the disturbance point to the instability limit point. , set the ratio threshold Set as Function of preset safety margin factor: ,in, This is the proportional threshold; This refers to the critical time for voltage instability. The safety margin factor is set to 0.85. The system sampling period is the average dwell time of control messages waiting to be sent in the queue. With physical validity window The ratio exceeds the proportion threshold. At that time, the control center reconstructs the full-precision numerical control command into a compressed control command vector with a length of 5 bits, where the first 2 bits are the adjustment direction sign bit and the last 3 bits are the sensitivity level bit.

[0043] The binary code 01 for the direction sign bit maps to increasing reactive power output, and the binary code 10 maps to decreasing reactive power output. The binary codes for the sensitivity level bits, from 000 to 111, correspond to eight preset sensitivity weights. The physical compensation step size corresponding to each level of binary code is determined by adjusting the local voltage and the reactive power sensitivity matrix. Logarithmic compression quantization is performed to obtain the voltage sample value from the previous time step. After the terminal receives the compression control command vector, it retrieves the voltage sample value from the register and matches the physical step size corresponding to the current binary code by looking up a table. The transient voltage compensation increment is then calculated and restored. The execution terminal superimposes the restored transient voltage compensation increment into the locally stored action history, driving the reactive power compensation unit to change its output power. This process eliminates the logical delay caused by full-precision data transmission through the bit mapping protocol, enabling the reactive power support action on the physical side to accurately align with the voltage transient even when the monitoring channel is congested. When the monitoring channel pressure index drops below 0.5, the system automatically switches to the local logic recovery mode, and performs closed-loop correction on the cumulative adjustment deviation generated in the compression mode through full-precision numerical control commands, so that the bus voltage quickly converges to the steady-state range near 1.0 pu, eliminating voltage oscillations induced by information uncertainty.

[0044] Example 5: In the field commissioning of a distribution network containing heterogeneous communication media, the smart grid monitoring system initiates a preset calibration program for the power network topology to which the controlled node belongs. It uses the Newton-Raphson method to calculate the Jacobian matrix on the primary side of the system, extracts the partial derivative elements of reactive power with respect to node voltage from the Jacobian matrix, and constructs an initial local voltage and reactive power sensitivity matrix based on the obtained partial derivative elements. The least squares fitting algorithm is used to combine the local voltage with the reactive power sensitivity matrix. The elements in the system are converted into preset local sensitivity constants within the execution terminal. The system sends 1,000 sets of probe messages to the controlled node in a undisturbed state, and statistically records the baseline round-trip delay and baseline packet loss rate. Based on the variance distribution of the baseline parameters, the weight coefficients of the packet loss rate variable, the round-trip delay variable, and the delay jitter variance variable in the weighted evaluation model are determined, so that the calculation method of the monitoring channel pressure index is statically aligned with the inherent noise characteristics of the specific communication physical link.

[0045] When the system faces changes in primary load model parameters or significant fluctuations in distributed power supply capacity, the monitoring system triggers dynamic baseline update logic. It periodically generates sensitivity correction coefficients corresponding to different topology states using an offline simulation platform and asynchronously writes these coefficients into the non-volatile register of the execution terminal via low-bandwidth priority management messages. This ensures that the transient voltage compensation increment is restored when the instruction dimensionality reduction and reconstruction conditions are met. To ensure physical consistency, the terminal periodically monitors the clock deviation between its local clock and the control center's master station. If the clock deviation exceeds 1ms, a nanosecond-level time synchronization compensation action is initiated to ensure physical consistency. The judgment criteria are kept consistent among the spatially distributed terminals, and the bus voltage fluctuations converge to the range of 0.98pu to 1.02pu within the preset 5 sampling periods.

[0046] Example 6: In a heterogeneous smart grid environment containing a fiber optic backbone and carrier communication branches, the monitoring system performs an offline sensitivity mapping calibration program for the execution terminal. It constructs a node impedance matrix using power network topology data, extracts feature vectors related to bus voltage stability, sets the sampling window length for the physical evolution rate to 20ms, and uses the Monte Carlo algorithm to simulate the system response trajectory under 10,000 sets of random load fluctuation sequences. It statistically analyzes the slope distribution of voltage response to reactive power disturbances at different nodes, and correlates the local voltage with the reactive power sensitivity matrix based on the probability density function of the response slope. The continuous parameters are divided into 8 equal-width intervals. The center value of each interval is mapped to the sensitivity level bit in the corresponding compression control command vector to obtain the bit mapping lookup table of the sensitivity level bit binary code from 000 to 111, and then written into the register of the execution terminal.

[0047] The system considers the safety margin coefficient in the instruction dimensionality reduction and reconstruction determination. An optimization parameter-finding program based on statistical probability is executed, which defines the recovery time of the bus voltage returning to the safe range as the objective function, and changes the absolute value of the voltage deviation in the offline simulation environment. The initial disturbance was determined, and the system stability performance was observed under different safety margin coefficients. When the value is in the range of 0.8 to 0.9, the recovery time of the bus voltage fluctuation is related to the physical effectiveness window. To achieve matching, when the safety margin coefficient When the value is greater than 0.95, the lenient judgment criteria result in a higher average dwell time for control messages awaiting delivery in the queue. Exceeding the physical validity window This can trigger oscillations, and in this embodiment, 0.85 is selected as the safety margin coefficient. The calibration values ​​are determined, and the weighted evaluation model is initialized using the offline extracted communication link round-trip delay benchmark, so that the criterion parameter matching is completed before the system is put into operation.

[0048] In the startup process of a regional distribution network containing various reactive power compensation devices, the monitoring system performs a quantitative calibration procedure for the physical response characteristics of the execution terminals. It uses offline data to obtain the response sensitivity curves of the controlled node bus voltage to changes in reactive power, and then converts the sensitivity matrix... The parameters are mapped to the sensitivity level bits in the compression control command vector through logarithmic normalization. The physical compensation range is divided into 8 binary steps, with each step corresponding to a bit state of the sensitivity level bit. This establishes a physical mapping relationship between the command bit stream and the actual output of the reactive power compensation unit. The monitoring system targets the physical effectiveness window. Initiate a benchmark calibration procedure based on transient evolution rate, utilizing the equivalent impedance of the primary side of the power system. The physical time boundary for the bus voltage to evolve from the current state to the over-limit state is calculated based on the load mutation rate. When the monitored transient evolution rate meets the preset rate threshold judgment condition, the system retrieves the locally stored mapping relationship between transient evolution and latency tolerance, calculates the maximum allowable transmission latency of the control command under the current physical evolution trend, and thus determines the physical validity window. The numerical value is determined, and a deviation correction based on the measured transient response is initiated during the self-check phase of the system's operating cycle. Specifically, the physical validity window... The calculation is based on the remaining time for the controlled node voltage to evolve from its current value to the instability critical threshold. The monitoring system uses the measured node equivalent impedance Z combined with the detected load current abrupt change rate to calculate the voltage sag slope, T. max The value is equal to the voltage safety margin (i.e., the difference between the current voltage and the 0.9 pu drop limit) divided by the voltage drop slope. Through this prediction model, the system can quantitatively determine the maximum time boundary that the physical side allows the control command to reach, thereby providing a definite time reference for command dimensionality reduction and reconstruction.

[0049] The embodiments of this application have been described above with reference to the accompanying drawings. Unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other. This application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit of this application and the scope of protection of this invention, and all of these forms are within the protection scope of this application.

Claims

1. A power system voltage stability collaborative optimization control method, applied to a smart grid monitoring system, comprising the following steps: Step S1: Collect the bus voltage sampling sequence of the controlled node, calculate the differential slope of the voltage sampling values ​​at adjacent sampling times, and obtain the transient evolution rate that characterizes the urgency of the state transition of the controlled node. Step S2: Extract the packet loss rate, round-trip time, and queue delay from the monitoring channel of the smart grid monitoring system, and calculate the monitoring channel pressure index that characterizes the degree of congestion in the monitoring channel. Step S3: Establish a coupling criterion between transient evolution rate and monitoring channel pressure index. By comparing the monitoring channel pressure index with a preset pressure threshold and the transient evolution rate with a preset rate threshold, determine whether the current control loop meets the instruction dimensionality reduction and reconstruction conditions. Step S4: If it is determined that the instruction dimensionality reduction and reconstruction conditions are not met, then a full-precision numerical control instruction is sent to the execution terminal to which the controlled node belongs. Step S5: If the condition for instruction dimensionality reduction and reconstruction is met, the adjustment direction is determined based on the full-precision numerical control instruction, the physical validity window corresponding to the current control cycle is calculated, the full-precision numerical control instruction is reconstructed into a compressed control instruction vector containing the adjustment direction sign bit and the sensitivity level bit, and the compressed control instruction vector is transmitted based on a preemptive tag. The compressed control instruction vector reaches the execution terminal within the physical validity window. Step S6: The execution terminal receives the compression control command vector, retrieves the voltage sampling value of the previous moment cached in the local register, restores the transient voltage compensation increment by combining it with the preset local sensitivity constant, and superimposes the transient voltage compensation increment into the locally stored action quantity history to drive the reactive power compensation unit connected to the execution terminal to change the output power.

2. The power system voltage stability collaborative optimization control method according to claim 1, characterized in that, The process of calculating the monitoring channel pressure index in step S2 includes: step S21, constructing a weighted evaluation model that includes packet loss rate variables, round-trip delay variables, and delay jitter variance variables; step S22, inputting the real-time collected communication performance parameters into the weighted evaluation model to obtain the monitoring channel pressure index; wherein, if the monitoring channel pressure index continuously exceeds the preset pressure threshold for 3 sampling periods, a channel congestion trigger signal is generated.

3. The power system voltage stability collaborative optimization control method according to claim 1, characterized in that, The process of obtaining the transient evolution rate in step S1 includes: step S11, obtaining the absolute value of voltage deviation for 5 consecutive sampling points within a preset sliding time window; step S12, calculating the first time derivative of the absolute value of voltage deviation and generating the transient evolution rate by combining it with a preset time decay factor.

4. The power system voltage stability collaborative optimization control method according to claim 1, characterized in that, The coupling criteria in step S3 include: establishing a binary decision matrix consisting of channel pressure dimension and physical evolution dimension; when the monitoring channel pressure index indicates that the available bandwidth of the monitoring channel is less than 128kbps and the transient evolution rate exceeds the preset rate threshold, the collaborative optimization control method switches the control mode from the central station centralized calculation mode to the execution terminal local logic recovery mode.

5. The power system voltage stability collaborative optimization control method according to claim 1, characterized in that, The method for generating the compression control command vector includes: mapping the transient evolution rate to a preset quantization range to determine the corresponding sensitivity level; the sensitivity level bit adopts a binary encoding method to instruct the execution terminal to adjust the step amplitude of the reactive power output within the physical validity window.

6. The power system voltage stability collaborative optimization control method according to claim 1, characterized in that, In step S5, the preemptive transmission tag is used to guide the smart grid's communication gateway to identify the specific protocol header of the compressed control command vector. The communication gateway allocates a reserved transmission time slot for the compressed control command vector, and the compressed control command vector arrives at the execution terminal within a 20ms physical response window.

7. The power system voltage stability collaborative optimization control method according to claim 1, characterized in that, The method also includes step S7, whereby after the monitoring channel pressure index falls below the preset pressure threshold, the control center sends a full-precision numerical control command to the execution terminal to perform closed-loop correction on the cumulative adjustment deviation generated by the execution terminal in the local logic recovery mode.

8. The power system voltage stability collaborative optimization control method according to claim 1, characterized in that, The execution terminal is a remote terminal unit or intelligent inverter distributed in the distribution network; the local register adopts a ring queue structure to dynamically store the voltage measurement values ​​and compensation action history of the most recent 5 sampling cycles.

9. The power system voltage stability collaborative optimization control method according to claim 1, characterized in that, The collaborative optimization control method is applied to smart grids that integrate distributed power sources and nonlinear loads. By coupling the congestion status of the monitoring channel to the voltage control loop, it achieves collaborative alignment between the control command transmission timescale and the power system physical evolution timescale.

Images