Time synchronization control method and system for distributed power generation system

Abstract

The invention discloses a time synchronization control method and system for a distributed power generation system, and the method comprises the following steps: obtaining the local clock deviation of each distributed unit, carrying out the time service calibration, collecting the calibrated operation power data, and generating a local power parameter sequence with a time label. The method comprises the following steps of: receiving a message, packaging the message into a synchronous message, sending the synchronous message to a main control end, pre-distributing the message, acquiring response data of each unit, evaluating the message receiving integrity of each unit at the same time point, and determining a final distribution strategy according to the message receiving integrity. And distributing a synchronous message according to the strategy, and calculating a deviation value between the local power of each unit and the average power of the whole network. And generating a regulation and control signal of each unit according to the power deviation value, thereby realizing power regulation and optimization of the distributed power generation system. According to the invention, the precision of system time synchronization and the coordination control efficiency are effectively improved, and the stability and economy of system operation are enhanced.

Description

Technical Field

[0001] This invention relates to the field of power generation control technology, and in particular to a time synchronization control method and system for distributed power generation systems. Background Technology

[0002] Distributed generation systems, through the parallel operation of multiple units, can effectively improve power supply reliability, flexibility, and energy utilization efficiency. Traditional parallel control schemes mainly rely on wired communication methods (such as CAN bus, Ethernet, etc.) to transmit the power parameters of each unit. Although wired communication has certain advantages in transmission reliability, it has obvious limitations in practical applications: complex wiring, large amount of engineering work, especially in application scenarios where distributed units are scattered and installation locations are variable, resulting in high deployment costs and poor scalability, making it difficult to adapt to the needs of flexible expansion and dynamic adjustment.

[0003] To overcome the shortcomings of wired communication, parallel control schemes based on wireless communication have emerged in recent years. These schemes utilize wireless technologies such as Wi-Fi, ZigBee, LoRa, or 4G / 5G to transmit power parameters, effectively reducing wiring complexity and improving system deployment flexibility. However, existing wireless parallel control schemes have significant shortcomings in time synchronization. Due to the lack of a unified microsecond-level high-precision time reference, discrepancies exist between the local clocks of each distributed unit, resulting in inconsistent power parameter acquisition times and making accurate comparison impossible on the same time dimension.

[0004] Furthermore, existing technologies mostly employ simple broadcast or unicast mechanisms for the transmission and distribution of power parameters, failing to fully consider the differences in wireless reception performance among units and the timing consistency of message transmission. This can easily lead to incomplete data reception or significant delays in some units, further reducing the accuracy of power equalization adjustment and dynamic response performance.

[0005] Therefore, there is an urgent need in this field for a wireless parallel control method and system that has high-precision time synchronization capability, can dynamically optimize message distribution strategy according to wireless transmission characteristics, and combines intelligent control algorithm to achieve rapid power balancing, so as to solve the problems of clock deviation, unreliable transmission, and insufficient adjustment accuracy in the existing technology, and improve the operational stability and economy of distributed generation system under variable operating conditions. Summary of the Invention

[0006] To address at least one of the aforementioned technical problems, this invention proposes a time synchronization control method and system for distributed generation systems.

[0007] The first aspect of this invention provides a time synchronization control method for a distributed generation system, comprising: The system acquires local clock deviation data of generators in each distributed unit of the distributed generation system, performs time synchronization calibration on each distributed unit based on the local clock deviation data, collects the operating power data of each distributed unit after time synchronization calibration in real time, and generates a local power parameter sequence with time tags. The local power parameter sequence is encapsulated into a synchronization message, the synchronization message is transmitted to the master control terminal, and the synchronization message is pre-distributed to obtain the response data of each distributed unit to the pre-distributed synchronization message. The integrity of each distributed unit's reception of the synchronization message at the same point in time is determined based on the response data, and the distribution strategy of the synchronization message is determined based on the integrity. The synchronization message is distributed to each distributed unit according to the distribution strategy, and the power deviation between the local power parameter and the average power parameter of the whole network is calculated. Based on the power deviation value, a control signal is generated for each distributed unit, and the power regulation of the distributed generation system is optimized based on the control signal.

[0008] In this scheme, the process of acquiring the local clock deviation data of each distributed unit generator in the distributed generation system, performing time synchronization calibration on each distributed unit based on the local clock deviation data, collecting the operating power data of each distributed unit after time synchronization calibration in real time, and generating a local power parameter sequence with time tags is as follows: A high-precision time synchronization source sends time synchronization pulse signals to each distributed unit, collects the response delay time of the local clock of each distributed unit to the time synchronization pulse signals, and calculates the local clock deviation data of each distributed unit by combining the sending time of the time synchronization pulse signals. The clock deviation is smoothed by a moving average filtering algorithm based on the local clock deviation data to generate a clock calibration compensation value. The local clock module of each distributed unit is dynamically adjusted based on the clock calibration compensation value. After the time synchronization calibration is completed, the active power and reactive power parameters of each distributed unit generator are synchronously collected at a fixed sampling period through the power sensor built into the distributed unit, and a unique time tag is assigned to each power parameter in combination with the calibrated local clock. Based on time tags, active and reactive power parameters collected at the same time are associated into data groups and arranged in chronological order to construct a local power parameter sequence.

[0009] In this scheme, the process of encapsulating the local power parameter sequence into a synchronization message, transmitting the synchronization message to the master control terminal, and performing a pre-distribution operation on the synchronization message to obtain the response data of each distributed unit to the pre-distributed synchronization message is as follows: The local power parameter sequence is encapsulated into a structured synchronization message, which includes a message header, a timestamp field, a power parameter sequence payload, and a cyclic redundancy check code. Based on the synchronization message, determine the amount of data and data length that need to be transmitted to each distributed unit per unit time, and construct a test synchronization message with the same amount of data and data length; The test synchronization message is pre-distributed to all distributed units simultaneously via multicast. After receiving the test synchronization message, each distributed unit verifies the integrity of the message according to the cyclic redundancy check code. If the verification fails, a reception error event is recorded and a local retransmission request is triggered. If the verification is successful, the system receives response frames returned by each distributed unit, obtains its own device identifier, local timestamp of the message reception time, received signal strength indicator value, and message verification status, and constructs the response data for the pre-distributed synchronization message.

[0010] In this scheme, determining the integrity of each distributed unit's reception of synchronization messages at the same time point based on the response data, and determining the distribution strategy of synchronization messages based on the integrity, specifically involves: Based on the local timestamp of the message reception time and the received signal strength indication value in the response frames returned by each distributed unit, a two-dimensional feature matrix of time and signal strength is constructed. A hierarchical clustering algorithm is introduced to perform cluster analysis on the two-dimensional feature matrix. A distance threshold for the hierarchical clustering algorithm is set, and the Euclidean distance of each distributed unit in the feature space is calculated. Distributed units whose Euclidean distance is less than the distance threshold are aggregated into the same receiving performance cluster to obtain several receiving performance clusters. For each receiving performance cluster, the response delay time data of all distributed units within the cluster is extracted. A kernel density estimation algorithm is introduced to estimate the probability density of the response delay time data, and the probability density distribution curve of the response delay time is calculated. The representative delay time of the receiving performance cluster is determined based on the peak position of the probability density distribution curve. Based on the representative delay time, each receiving performance cluster is sorted in ascending order of delay time to construct an ordered chain of delay time. Based on the aforementioned delay time, the receiving performance cluster with the smallest delay time is selected as the benchmark receiving cluster. The representative delay time of the benchmark receiving cluster is obtained as the network-wide synchronization benchmark time. Based on the network-wide synchronization benchmark time, the relative time difference between other receiving performance clusters and the benchmark receiving cluster is calculated. An adaptive step size search algorithm is introduced, using the relative time difference as a constraint, to initialize the data block size trial value. Based on the representative delay time of each receiving performance cluster and the data block size trial value, the theoretical transmission time for each receiving performance cluster to complete data transmission at the network-wide synchronization reference time is calculated. It is then determined whether the theoretical transmission time of all receiving performance clusters meets the constraint of the network-wide synchronization reference time. If there are receiving performance clusters that do not meet the constraints, the trial value of the data block size is reduced according to the preset step size reduction rule and recalculated until the theoretical transmission time of all receiving performance clusters meets the constraints. The trial value of the data block size at this time is determined as the unified maximum synchronous data volume of the entire network. Based on the unified maximum synchronization data volume across the entire network, the synchronization messages to be distributed are divided into several data distribution blocks. For each data distribution block, the target distribution sequence of the data distribution block is determined based on the clustering results of the receiving performance cluster and the ordered chain of delay time. Based on the target distribution sequence, a distribution strategy for the synchronization messages is constructed. The distribution strategy includes the sending time of each data distribution block, the target receiving cluster, and the corresponding transmission channel parameters.

[0011] In this scheme, the step of distributing the synchronization message to each distributed unit according to the distribution strategy and calculating the power deviation between the local power parameter and the average power parameter of the entire network is as follows: According to the distribution strategy, each data distribution block of the synchronization message is distributed to each distributed unit in sequence through the corresponding communication link. After receiving the data distribution block, each distributed unit verifies and reassembles the data distribution block to restore the complete synchronization message and extract the whole network power parameter sequence from it. Obtain the local power parameter sequence of each distributed unit in the time period corresponding to the synchronization message, perform time alignment processing on the local power parameter sequence and the whole network power parameter sequence, and use the timestamp of the whole network power parameter sequence as the reference to resample the local power parameter sequence using cubic spline interpolation method to ensure that the power parameters of each distributed unit are strictly synchronized in time. Calculate the average power parameters of the entire network at each time point for the time-aligned power parameter sequence of the entire network. The average power parameters of the entire network include the average active power and the average reactive power of the entire network. The difference between the local active power of each distributed unit and the average active power of the entire network at each time point is calculated sequentially to obtain the active power deviation value at each time point. At the same time, the difference between the local reactive power of each distributed unit and the average reactive power of the entire network at each time point is calculated to obtain the reactive power deviation value at each time point. The active power deviation and reactive power deviation values ​​at all time points of each distributed unit are integrated to obtain the cumulative active power deviation and cumulative reactive power deviation of each distributed unit within the time period. The cumulative active power deviation and cumulative reactive power deviation are then output as the power deviation values ​​of each distributed unit.

[0012] In this scheme, the step of generating a control signal for each distributed unit based on the power deviation value, and optimizing the power regulation of the distributed generation system based on the control signal, specifically involves: Based on the power deviation value, construct the active power deviation sequence and reactive power deviation sequence of each distributed unit, introduce the fuzzy PID control algorithm, and set the initial values ​​of the proportional coefficient, integral coefficient and derivative coefficient of the fuzzy PID control algorithm. Calculate the rate of change of active power deviation and the rate of change of reactive power deviation based on the active power deviation sequence and the reactive power deviation sequence, respectively. The active power deviation value and reactive power deviation value are used as input variables of the fuzzy PID control algorithm. The proportional coefficient, integral coefficient and derivative coefficient are adaptively adjusted through the fuzzy rule base to obtain the optimized proportional coefficient, integral coefficient and derivative coefficient. The active power control signal and reactive power control signal of each distributed unit are calculated based on the optimized proportional coefficient, integral coefficient, and derivative coefficient, respectively. The active power control signal is a speed control signal, and the reactive power control signal is a voltage control signal. The speed control signal is output to the speed governor of each distributed unit, and the active power output of the generator is changed by adjusting the fuel supply or steam intake of the prime mover. The voltage control signal is output to the automatic voltage regulator of each distributed unit, and the terminal voltage and reactive power output of the generator are changed by adjusting the excitation current. The active and reactive power parameters of each distributed unit after regulation are monitored in real time, and the power deviation value is recalculated. If the power deviation value does not reach the preset convergence threshold, the fuzzy PID regulation process is repeated until the power deviation value of each distributed unit is less than the convergence threshold.

[0013] A second aspect of the present invention also provides a time synchronization control system for a distributed generation system, the system comprising: a memory and a processor, wherein the memory includes a time synchronization control method program for the distributed generation system, and when the time synchronization control method program for the distributed generation system is executed by the processor, the following steps are implemented: The system acquires local clock deviation data of generators in each distributed unit of the distributed generation system, performs time synchronization calibration on each distributed unit based on the local clock deviation data, collects the operating power data of each distributed unit after time synchronization calibration in real time, and generates a local power parameter sequence with time tags. The local power parameter sequence is encapsulated into a synchronization message, the synchronization message is transmitted to the master control terminal, and the synchronization message is pre-distributed to obtain the response data of each distributed unit to the pre-distributed synchronization message. The integrity of each distributed unit's reception of the synchronization message at the same point in time is determined based on the response data, and the distribution strategy of the synchronization message is determined based on the integrity. The synchronization message is distributed to each distributed unit according to the distribution strategy, and the power deviation between the local power parameter and the average power parameter of the whole network is calculated. Based on the power deviation value, a control signal is generated for each distributed unit, and the power regulation of the distributed generation system is optimized based on the control signal.

[0014] A time synchronization control method and system for distributed generation systems includes the following steps: acquiring the local clock deviation of each distributed unit and performing time synchronization calibration; collecting calibrated operating power data; and generating a time-stamped sequence of local power parameters. These parameters are then encapsulated into a synchronization message and sent to the master control unit. The message is pre-distributed, and the response data of each unit is acquired to evaluate the message reception integrity of each unit at the same time point, thereby determining the final distribution strategy. The synchronization message is distributed according to this strategy, and the deviation between the local power of each unit and the average power of the entire network is calculated. Based on this power deviation, control signals for each unit are generated to optimize the power regulation of the distributed generation system. This invention effectively improves the accuracy of system time synchronization and the efficiency of coordinated control, enhancing the stability and economy of system operation. Attached Figure Description

[0015] Figure 1 A flowchart of a time synchronization control method for a distributed generation system according to the present invention is shown; Figure 2 A flowchart illustrating the present invention for obtaining response data from each distributed unit is shown; Figure 3 A flowchart illustrating the calculation of the power deviation between local power parameters and the average power parameters of the entire network is shown. Figure 4 A block diagram of a time synchronization control system for a distributed generation system according to the present invention is shown. Detailed Implementation

[0016] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0017] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0018] Figure 1 A flowchart of a time synchronization control method for a distributed generation system according to the present invention is shown.

[0019] like Figure 1 As shown, the first aspect of the present invention provides a time synchronization control method for a distributed generation system, comprising: S102, acquire the local clock deviation data of each distributed unit generator in the distributed generation system, perform time synchronization calibration on each distributed unit according to the local clock deviation data, collect the operating power data of each distributed unit after time synchronization calibration in real time, and generate a local power parameter sequence with time tag. S104, the local power parameter sequence is encapsulated into a synchronization message, the synchronization message is transmitted to the master control terminal, and the synchronization message is pre-distributed to obtain the response data of each distributed unit to the pre-distributed synchronization message; S106, determine the integrity of each distributed unit's reception of the synchronization message at the same time point based on the response data, and determine the distribution strategy of the synchronization message based on the integrity. S108, Distribute the synchronization message to each distributed unit according to the distribution strategy, and calculate the power deviation between the local power parameter and the average power parameter of the whole network; S110, generate a control signal for each distributed unit based on the power deviation value, and optimize the power regulation of the distributed generation system based on the control signal.

[0020] It should be noted that by using precise time synchronization and local clock calibration, the time base of the entire system is unified, eliminating the inherent timing errors caused by hardware crystal oscillator drift. Next, local data with time stamps is encapsulated and pre-distributed, actively probing the actual network environment to identify non-uniform communication delays and potential packet loss risks caused by differences in transmission paths and node processing capabilities. Based on this, an intelligent distribution strategy is formulated to dynamically adapt to network conditions. Through data segmentation and scheduling, it is ensured that even the slowest node receives a complete global data view before the system's agreed-upon logical time, thus creating conditions for true synchronous computation. The power deviation calculation performed on this basis, since all data is strictly aligned in time, accurately reflects the instantaneous differences between each unit and the system average state, rather than spurious deviations caused by timing misalignments. Finally, the control signal generated based on this precise deviation can drive the actuators to perform fine adjustments, guiding the system power distribution to converge quickly and smoothly to the desired equilibrium state.

[0021] According to an embodiment of the present invention, the step of acquiring local clock deviation data of each distributed unit generator in the distributed generation system, performing time synchronization calibration on each distributed unit based on the local clock deviation data, collecting the operating power data of each distributed unit after time synchronization calibration in real time, and generating a local power parameter sequence with time tags specifically includes: A high-precision time synchronization source sends time synchronization pulse signals to each distributed unit, collects the response delay time of the local clock of each distributed unit to the time synchronization pulse signals, and calculates the local clock deviation data of each distributed unit by combining the sending time of the time synchronization pulse signals. The clock deviation is smoothed by a moving average filtering algorithm based on the local clock deviation data to generate a clock calibration compensation value. The local clock module of each distributed unit is dynamically adjusted based on the clock calibration compensation value. After the time synchronization calibration is completed, the active power and reactive power parameters of each distributed unit generator are synchronously collected at a fixed sampling period through the power sensor built into the distributed unit, and a unique time tag is assigned to each power parameter in combination with the calibrated local clock. Based on time tags, active and reactive power parameters collected at the same time are associated into data groups and arranged in chronological order to construct a local power parameter sequence.

[0022] It should be noted that by relying on a high-precision time source to directly measure and calculate the absolute clock deviation of each distributed unit, the physical error between its internal clock and standard time is identified and quantified at its source. Then, a moving average filtering algorithm is used to smooth the raw deviation data, effectively filtering out noise caused by signal jitter or transient interference, generating a stable and reliable clock calibration compensation value. This enables dynamic fine-tuning of the local clock module, ensuring its long-term timing accuracy and stability.

[0023] Figure 2 A flowchart illustrating the present invention for obtaining response data for each distributed unit is shown.

[0024] According to an embodiment of the present invention, the step of encapsulating the local power parameter sequence into a synchronization message, transmitting the synchronization message to the master control terminal, and performing a pre-distribution operation on the synchronization message to obtain the response data of each distributed unit to the pre-distributed synchronization message specifically includes: The local power parameter sequence is encapsulated into a structured synchronization message, which includes a message header, a timestamp field, a power parameter sequence payload, and a cyclic redundancy check code. Based on the synchronization message, determine the amount of data and data length that need to be transmitted to each distributed unit per unit time, and construct a test synchronization message with the same amount of data and data length; The test synchronization message is pre-distributed to all distributed units simultaneously via multicast. After receiving the test synchronization message, each distributed unit verifies the integrity of the message according to the cyclic redundancy check code. If the verification fails, a reception error event is recorded and a local retransmission request is triggered. If the verification is successful, the system receives response frames returned by each distributed unit, obtains its own device identifier, local timestamp of the message reception time, received signal strength indicator value, and message verification status, and constructs the response data for the pre-distributed synchronization message.

[0025] It should be noted that in real-world network environments, the communication link quality, signal strength, and processing latency of each distributed unit vary. If critical synchronization data is distributed directly without prior investigation, some nodes may fail to synchronize on the same timeframe due to packet loss or excessive latency. Therefore, before actually performing power synchronization via synchronization messages, a comprehensive and realistic "stress test" and "performance probe" are conducted on the current network communication environment and the receiving capabilities of each node. By sending test messages that are isomorphic to real data, the actual load of the synchronization process can be simulated, thereby accurately measuring the actual transmission latency experienced by the messages when they reach different nodes, assessing signal transmission quality, and exposing potential packet loss or data corruption risks in advance. The collected response data, including precise reception time, signal strength, and verification status, provides the master control unit with a detailed real-time map of network performance.

[0026] According to an embodiment of the present invention, the step of determining the reception integrity of synchronization messages by each distributed unit at the same time point based on the response data, and determining the distribution strategy of synchronization messages based on the integrity, specifically includes: Based on the local timestamp of the message reception time and the received signal strength indication value in the response frames returned by each distributed unit, a two-dimensional feature matrix of time and signal strength is constructed. A hierarchical clustering algorithm is introduced to perform cluster analysis on the two-dimensional feature matrix. A distance threshold for the hierarchical clustering algorithm is set, and the Euclidean distance of each distributed unit in the feature space is calculated. Distributed units whose Euclidean distance is less than the distance threshold are aggregated into the same receiving performance cluster to obtain several receiving performance clusters. For each receiving performance cluster, the response delay time data of all distributed units within the cluster is extracted. A kernel density estimation algorithm is introduced to estimate the probability density of the response delay time data, and the probability density distribution curve of the response delay time is calculated. The representative delay time of the receiving performance cluster is determined based on the peak position of the probability density distribution curve. Based on the representative delay time, each receiving performance cluster is sorted in ascending order of delay time to construct an ordered chain of delay time. Based on the aforementioned delay time, the receiving performance cluster with the smallest delay time is selected as the benchmark receiving cluster. The representative delay time of the benchmark receiving cluster is obtained as the network-wide synchronization benchmark time. Based on the network-wide synchronization benchmark time, the relative time difference between other receiving performance clusters and the benchmark receiving cluster is calculated. An adaptive step size search algorithm is introduced, using the relative time difference as a constraint, to initialize the data block size trial value. Based on the representative delay time of each receiving performance cluster and the data block size trial value, the theoretical transmission time for each receiving performance cluster to complete data transmission at the network-wide synchronization reference time is calculated. It is then determined whether the theoretical transmission time of all receiving performance clusters meets the constraint of the network-wide synchronization reference time. If there are receiving performance clusters that do not meet the constraints, the trial value of the data block size is reduced according to the preset step size reduction rule and recalculated until the theoretical transmission time of all receiving performance clusters meets the constraints. The trial value of the data block size at this time is determined as the unified maximum synchronous data volume of the entire network. Based on the unified maximum synchronization data volume across the entire network, the synchronization messages to be distributed are divided into several data distribution blocks. For each data distribution block, the target distribution sequence of the data distribution block is determined based on the clustering results of the receiving performance cluster and the ordered chain of delay time. Based on the target distribution sequence, a distribution strategy for the synchronization messages is constructed. The distribution strategy includes the sending time of each data distribution block, the target receiving cluster, and the corresponding transmission channel parameters.

[0027] It should be noted that due to differences in communication link quality, signal transmission distance, and the degree of electromagnetic interference in the field, the data reception performance reflected by the local timestamp and received signal strength indication value in the response frame exhibits high dispersion. If a fixed amount of data is directly used for synchronization message distribution, some distributed units with poor reception performance will not have completed data reception at the network-wide synchronization reference time, while some distributed units with better reception performance will have completed reception in advance and are waiting for the next time sequence. This will cause the entire network to be unable to form a complete and consistent power parameter dataset at the same time point, resulting in data loss or misalignment in the time dimension for subsequent power deviation calculation and control signal generation. This will cause the distributed generation system to lose the network-wide synchronization reference for power regulation optimization, leading to technical problems such as inconsistent power regulation times among distributed units, grid-connected power fluctuations, and even a decrease in system stability. By constructing a two-dimensional time-signal strength characteristic... The system employs a hierarchical clustering algorithm to analyze the receiving performance of each distributed unit, forming clusters with similar communication characteristics. Based on this, a kernel density estimation algorithm is used to extract representative delay times from each cluster and construct an ordered chain of delay times. The cluster with the smallest delay time is used as the network-wide synchronization reference time. An adaptive step-size search algorithm iteratively optimizes and determines the maximum unified synchronization data volume that satisfies the time constraints of all receiving performance clusters. Finally, based on the clustering results and the ordered chain of delay times, a distribution strategy is constructed that includes transmission time, target receiving cluster, and transmission channel parameters. This achieves the goal of ensuring that all distributed units in the network receive the same amount of synchronization messages at the same time under different communication performance conditions. It solves the problem of asynchronous power regulation command execution caused by time synchronization deviation in the power control of distributed generation systems, significantly improving the coordination and stability of network-wide power regulation and ensuring the grid-friendliness and power quality of distributed generation systems.

[0028] Figure 3 The flowchart illustrating the calculation of the power deviation between local power parameters and the average power parameters of the entire network is shown.

[0029] According to an embodiment of the present invention, the step of distributing the synchronization message to each distributed unit according to the distribution strategy and calculating the power deviation between the local power parameter and the average power parameter of the entire network specifically involves: According to the distribution strategy, each data distribution block of the synchronization message is distributed to each distributed unit in sequence through the corresponding communication link. After receiving the data distribution block, each distributed unit verifies and reassembles the data distribution block to restore the complete synchronization message and extract the whole network power parameter sequence from it. Obtain the local power parameter sequence of each distributed unit in the time period corresponding to the synchronization message, perform time alignment processing on the local power parameter sequence and the whole network power parameter sequence, and use the timestamp of the whole network power parameter sequence as the reference to resample the local power parameter sequence using cubic spline interpolation method to ensure that the power parameters of each distributed unit are strictly synchronized in time. Calculate the average power parameters of the entire network at each time point for the time-aligned power parameter sequence of the entire network. The average power parameters of the entire network include the average active power and the average reactive power of the entire network. The difference between the local active power of each distributed unit and the average active power of the entire network at each time point is calculated sequentially to obtain the active power deviation value at each time point. At the same time, the difference between the local reactive power of each distributed unit and the average reactive power of the entire network at each time point is calculated to obtain the reactive power deviation value at each time point. The active power deviation and reactive power deviation values ​​at all time points of each distributed unit are integrated to obtain the cumulative active power deviation and cumulative reactive power deviation of each distributed unit within the time period. The cumulative active power deviation and cumulative reactive power deviation are then output as the power deviation values ​​of each distributed unit.

[0030] It's important to note that in a parallel system, the ideal state is for all units to share the total load equally, meaning each unit's output should match the average output of all units in the network. By calculating the deviation between the unit's power and this dynamically changing average power, the system can clearly identify which units are currently outputting too much (positive deviation) and which are outputting too little (negative deviation). This deviation is directly translated into the strength and direction of the control signal: a positive deviation requires reducing the unit's output, while a negative deviation requires increasing its output. Therefore, this calculation step is crucial in transforming the macroscopic goal of "equal power distribution" into specific, closed-loop negative feedback control commands that each distributed unit can execute. It serves as the decision-making basis for driving the entire system to converge from its current unbalanced state towards the target equilibrium state.

[0031] According to an embodiment of the present invention, the step of generating a control signal for each distributed unit based on the power deviation value, and optimizing the power regulation of the distributed generation system based on the control signal, specifically includes: Based on the power deviation value, construct the active power deviation sequence and reactive power deviation sequence of each distributed unit, introduce the fuzzy PID control algorithm, and set the initial values ​​of the proportional coefficient, integral coefficient and derivative coefficient of the fuzzy PID control algorithm. Calculate the rate of change of active power deviation and the rate of change of reactive power deviation based on the active power deviation sequence and the reactive power deviation sequence, respectively. The active power deviation value and reactive power deviation value are used as input variables of the fuzzy PID control algorithm. The proportional coefficient, integral coefficient and derivative coefficient are adaptively adjusted through the fuzzy rule base to obtain the optimized proportional coefficient, integral coefficient and derivative coefficient. The active power control signal and reactive power control signal of each distributed unit are calculated based on the optimized proportional coefficient, integral coefficient, and derivative coefficient, respectively. The active power control signal is a speed control signal, and the reactive power control signal is a voltage control signal. The speed control signal is output to the speed governor of each distributed unit, and the active power output of the generator is changed by adjusting the fuel supply or steam intake of the prime mover. The voltage control signal is output to the automatic voltage regulator of each distributed unit, and the terminal voltage and reactive power output of the generator are changed by adjusting the excitation current. The active and reactive power parameters of each distributed unit after regulation are monitored in real time, and the power deviation value is recalculated. If the power deviation value does not reach the preset convergence threshold, the fuzzy PID regulation process is repeated until the power deviation value of each distributed unit is less than the convergence threshold.

[0032] It should be noted that by introducing a fuzzy PID control algorithm, an intelligent closed-loop regulation mechanism capable of adapting to the nonlinear and time-varying characteristics of the system is constructed. The fuzzy rule base is the brain of this mechanism; it is an expert knowledge base containing a series of "IF-THEN" type linguistic rules. Its content mainly covers two aspects: first, evaluation rules for the magnitude of power deviation (such as "positive large," "positive small," "zero," "negative small," "negative large"); and second, evaluation rules for the rate of change of power deviation (such as "rapid increase," "slow decrease," etc.). The system fuzzifies the precise power deviation value and its rate of change into fuzzy quantities, and then uses the rule base for inference to dynamically and nonlinearly adjust the proportional, integral, and derivative coefficients of the PID controller. For example, when the deviation is large, the rule base will output an instruction to increase the proportional coefficient to speed up the response; when the deviation is close to zero but still shows a trend of change, it will adjust the derivative coefficient to suppress overshoot. This rule-based adaptive adjustment enables the generated active power speed regulation signal and reactive power voltage regulation signal to not only respond quickly, but also drive the speed governor and automatic voltage regulator (AVR) smoothly and accurately. Ultimately, it guides the power output of each distributed unit to converge quickly and smoothly to the target state of average distribution. This effectively avoids the oscillation or slow regulation problems that are prone to occur in traditional PID control under complex operating conditions, and significantly improves the stability, balance and dynamic response quality of the entire grid-connected system.

[0033] According to an embodiment of the present invention, it further includes: Based on the unified power control signal issued by the main control terminal, the physical response delay time of each distributed unit from receiving the control signal to the actual change of its power generation parameters is obtained. Based on the physical response delay time, a hierarchical clustering algorithm is used to aggregate distributed units with similar response characteristics into several device response clusters, and the average response delay time and the maximum delay difference within each device response cluster are calculated. Based on the average response delay time and the maximum delay difference within the cluster, a dedicated delay compensation time window is constructed for each device response cluster. The start time of the delay compensation time window is set according to the difference between the average response delay time of the cluster and the average response delay time of the fastest response cluster, and the length of the time window is adaptively determined according to the maximum delay difference within the cluster. Based on the delay compensation time window constructed for each device response cluster, the unified control signal is reconstructed in time sequence to generate a scheduling instruction sequence with differentiated issuance times; During the refactoring process, the scheduling instructions for the slowest responding device cluster are issued earliest, so that the start time of its execution can be aligned with that of other clusters. After the scheduling instruction sequence is issued, the actual power output curve of each distributed unit within the corresponding delay compensation time window is collected in real time and compared with the ideal power output change curve generated based on the unified control signal to calculate the actual power following error curve of each unit within the time window. Based on the actual power following error curve, a rolling optimization algorithm is introduced to fine-tune the start time and window length of the delay compensation time window for the corresponding device response cluster, so as to minimize the area enclosed by the actual power following error curve and the zero error baseline, thereby dynamically optimizing the delay compensation strategy of each cluster and realizing the time synchronization of the entire network power regulation at the physical execution level.

[0034] According to an embodiment of the present invention, the real-time acquisition of the actual power output curve of each distributed unit within the corresponding delay compensation time window, and the comparison with the ideal power output change curve generated based on the unified control signal, to calculate the actual power following error curve of each unit within the time window, specifically: Within the delay compensation time window, the active and reactive power output values ​​of each distributed unit are synchronously collected at a sampling rate higher than the power regulation dynamic process frequency, and the actual power-time series of each unit is formed based on the unified timestamp after time synchronization calibration. Based on the target value of the unified control signal and its theoretical change model, an ideal power output change curve from the control start point to the target value is generated, and this curve is discretized according to the same timestamp as the actual power collected to form an ideal power-time series. By subtracting the actual power-time series data points from the ideal power-time series data points at the same time and in the same distributed unit, the instantaneous error values ​​of active power and reactive power at each sampling time are calculated respectively. By sequentially connecting the instantaneous error values ​​at each sampling moment along the time axis, the actual power following error curve of the distributed unit within the delay compensation time window of this control action is formed. The actual power following error curve fully characterizes the dynamic deviation trajectory between the actual output and the theoretical expectation of the unit from the start of response to the process of basically reaching a stable state.

[0035] It should be noted that in real-world scenarios where distributed generation systems contain multiple models of older equipment, the inherent differences in the processing speed of controllers and the mechanical and electrical inertia of actuators from different manufacturing eras and manufacturers lead to significant physical response asynchrony when receiving and executing a unified power regulation signal issued at the same time. This hardware-level response delay difference causes theoretically precise and coordinated control strategies to become misaligned in actual execution. The changes in power output of each unit cannot be aligned on the time axis, which not only fails to achieve the expected power balance regulation effect but may also trigger new power oscillations due to inconsistent actions, severely restricting the accuracy and stability of the coordinated control of the distributed generation system.

[0036] Therefore, this invention actively measures and clusters the actual physical response delay of each unit, and accordingly tailors differentiated instruction issuance times for device clusters with different response speeds, thus constructing a delay compensation time window. This aligns asynchronous response processes to the same starting point at the source. The technical effect is that this measurement-based active compensation mechanism ensures that regardless of the device's response speed, the start time of its execution is synchronized at the physical time level, eliminating the initial action time difference caused by hardware differences at its root. Furthermore, by comparing and analyzing the dynamic deviation between the compensated actual output and the ideal curve, and using a rolling optimization algorithm to fine-tune the compensation parameters online, this method continuously optimizes the compensation effect, adaptively converges the actual power output trajectory of each unit, and ultimately makes the power adjustment actions of all distributed units in the network highly coordinated in time, with output curves becoming more consistent in shape. This significantly improves the accuracy, speed, and stability of system-level power control, effectively suppresses power oscillations caused by asynchronous execution, and ensures high-quality grid-connected operation of the distributed generation system.

[0037] Figure 4 A block diagram of a time synchronization control system for a distributed generation system according to the present invention is shown.

[0038] A second aspect of the present invention also provides a time synchronization control system for a distributed generation system. The system includes a memory 401, a processor 402, and a communication interface 403. The memory includes a time synchronization control method program for the distributed generation system. The communication interface is used for data connection and communication between the memory and the processor. When the processor executes the time synchronization control method program for the distributed generation system, it performs the following steps: The system acquires local clock deviation data of generators in each distributed unit of the distributed generation system, performs time synchronization calibration on each distributed unit based on the local clock deviation data, collects the operating power data of each distributed unit after time synchronization calibration in real time, and generates a local power parameter sequence with time tags. The local power parameter sequence is encapsulated into a synchronization message, the synchronization message is transmitted to the master control terminal, and the synchronization message is pre-distributed to obtain the response data of each distributed unit to the pre-distributed synchronization message. The integrity of each distributed unit's reception of the synchronization message at the same point in time is determined based on the response data, and the distribution strategy of the synchronization message is determined based on the integrity. The synchronization message is distributed to each distributed unit according to the distribution strategy, and the power deviation between the local power parameter and the average power parameter of the whole network is calculated. Based on the power deviation value, a control signal is generated for each distributed unit, and the power regulation of the distributed generation system is optimized based on the control signal.

[0039] A time synchronization control method and system for distributed generation systems includes the following steps: acquiring the local clock deviation of each distributed unit and performing time synchronization calibration; collecting calibrated operating power data; and generating a time-stamped sequence of local power parameters. These parameters are then encapsulated into a synchronization message and sent to the master control unit. The message is pre-distributed, and the response data of each unit is acquired to evaluate the message reception integrity of each unit at the same time point, thereby determining the final distribution strategy. The synchronization message is distributed according to this strategy, and the deviation between the local power of each unit and the average power of the entire network is calculated. Based on this power deviation, control signals for each unit are generated to optimize the power regulation of the distributed generation system. This invention effectively improves the accuracy of system time synchronization and the efficiency of coordinated control, enhancing the stability and economy of system operation.

[0040] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.

[0041] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units. They may be located in one place or distributed across multiple network units. Some or all of the units may be selected to achieve the purpose of this embodiment according to actual needs.

[0042] In addition, in the various embodiments of the present invention, each functional unit can be integrated into one processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit; the integrated unit can be implemented in hardware or in the form of hardware plus software functional units.

[0043] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0044] Alternatively, if the integrated units of this invention are implemented as software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of this invention, or the parts that contribute to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROM, RAM, magnetic disks, or optical disks.

[0045] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A time synchronization control method for a distributed power generation system, characterized by, The method comprises the following steps: obtaining local clock deviation data of each distributed unit generator in the distributed power generation system, calibrating the time of each distributed unit according to the local clock deviation data, collecting the running power data of each distributed unit after the time calibration in real time, and generating a local power parameter sequence with a time label; encapsulating the local power parameter sequence into a synchronization message, transmitting the synchronization message to the master control end, and performing a pre-distribution operation on the synchronization message to obtain the response data of each distributed unit to the pre-distributed synchronization message; determining the receiving integrity of each distributed unit to the synchronization message at the same time point according to the response data, and determining the distribution strategy of the synchronization message according to the integrity; distributing the synchronization message to each distributed unit according to the distribution strategy, and calculating the power deviation value of the local power parameter and the global average power parameter; generating a control signal for each distributed unit according to the power deviation value, and adjusting and optimizing the power of the distributed power generation system according to the control signal.

2. The time synchronization control method for a distributed power generation system according to claim 1, wherein The method comprises the following steps: sending a time calibration pulse signal to each distributed unit based on a high-precision time source, collecting the response delay time of the local clock of each distributed unit to the time calibration pulse signal, and calculating the local clock deviation data of each distributed unit in combination with the sending time of the time calibration pulse signal; smoothing the clock deviation by using a sliding average filtering algorithm according to the local clock deviation data, generating a clock calibration compensation value, and dynamically adjusting the local clock module of each distributed unit through the clock calibration compensation value; after the time calibration is completed, synchronously collecting the active power and reactive power parameters of each distributed unit generator by the power sensor built in the distributed unit at a fixed sampling period, and assigning a unique time label to each power parameter in combination with the calibrated local clock; based on the time label, associating the active power and reactive power parameters collected at the same time into a data group, arranging and constructing a local power parameter sequence in chronological order.

3. The time synchronization control method for a distributed power generation system according to claim 1, wherein The method comprises the following steps: encapsulating the local power parameter sequence into a structured synchronization message, wherein the structured synchronization message comprises a message header, a timestamp field, a power parameter sequence load, and a cyclic redundancy check code; determining the data amount and data length that need to be transmitted to each distributed unit per unit time according to the synchronization message, and constructing a test synchronization message with the same data amount and data length as the synchronization message; The test synchronization message is pre-distributed to all distributed units in a multicast manner, and each distributed unit verifies the integrity of the message according to a cyclic redundancy check code after receiving the test synchronization message, and records a receiving error event and triggers a local retransmission request if the verification fails; If the verification is successful, the response frame returned by each distributed unit is received, the self-device identifier, the local timestamp of the message receiving time, the received signal strength indication value and the message verification state of the response frame are obtained, and the response data of the pre-distribution synchronization message is constituted.

4. The time synchronization control method for a distributed power generation system according to claim 1, characterized by, The receiving integrity of each distributed unit to the synchronization message at the same time point is determined according to the response data, and the distribution strategy of the synchronization message is determined according to the integrity, specifically: A time-signal strength two-dimensional feature matrix is constructed based on the local timestamp of the message receiving time and the received signal strength indication value in the response frame returned by each distributed unit, a hierarchical clustering algorithm is introduced to perform clustering analysis on the two-dimensional feature matrix, a distance threshold of the hierarchical clustering algorithm is set, the Euclidean distance of each distributed unit in the feature space is calculated, the distributed units with the Euclidean distance less than the distance threshold are aggregated into the same receiving performance cluster, and a plurality of receiving performance clusters are obtained; For each receiving performance cluster, the response delay time data of all distributed units in the cluster is extracted, a kernel density estimation algorithm is introduced to perform probability density estimation on the response delay time data, a probability density distribution curve of the response delay time is calculated, a representative delay time of the receiving performance cluster is determined according to the peak position of the probability density distribution curve, and each receiving performance cluster is sorted according to the delay time from small to large according to the representative delay time, and a delay time ordered chain is constructed; According to the delay time, the receiving performance cluster with the smallest delay time is taken as a reference receiving cluster, the representative delay time of the reference receiving cluster is taken as a network-wide synchronization reference time, and the relative time difference value of other receiving performance clusters and the reference receiving cluster is calculated based on the network-wide synchronization reference time; An adaptive step search algorithm is introduced, the relative time difference value is taken as a constraint condition, a data block size tentative value is initialized, the theoretical transmission time length of each receiving performance cluster to complete data transmission at the network-wide synchronization reference time point is calculated according to the representative delay time of each receiving performance cluster and the data block size tentative value, and it is judged whether the theoretical transmission time length of all receiving performance clusters satisfies the constraint of the network-wide synchronization reference time. If there is a receiving performance cluster that does not satisfy the constraint, the data block size tentative value is reduced according to a preset step reduction rule and recalculated until the theoretical transmission time length of all receiving performance clusters satisfies the constraint condition, and the data block size tentative value at this time is determined as the network-wide unified maximum synchronization data amount. According to the maximum synchronization data amount of the whole network, the synchronization message to be distributed is divided into a plurality of data distribution blocks, for each data distribution block, a target distribution time sequence of the data distribution block is determined according to the clustering result and the delay time ordered chain of the receiving performance cluster, a distribution strategy of the synchronization message is constructed according to the target distribution time sequence, and the distribution strategy includes a sending time of each data distribution block, a target receiving cluster and corresponding transmission channel parameters.

5. The time synchronization control method for a distributed power generation system according to claim 1, characterized by, According to the distribution strategy, the synchronization message is distributed to each distributed unit, and a power deviation value of a local power parameter and a whole network average power parameter is calculated, specifically as follows: According to the distribution strategy, each data distribution block of the synchronization message is sequentially distributed to each distributed unit through a corresponding communication link, each distributed unit checks and reorganizes the data distribution block after receiving the data distribution block, restores a complete synchronization message and extracts a whole network power parameter sequence in the synchronization message; A local power parameter sequence of each distributed unit in a time period corresponding to the synchronization message is obtained, the local power parameter sequence and the whole network power parameter sequence are time-aligned, a timestamp of the whole network power parameter sequence is taken as a reference, a cubic spline interpolation method is used to resample the local power parameter sequence, so that the power parameters of each distributed unit are strictly synchronized in time; A whole network average power parameter at each time point is calculated for the time-aligned whole network power parameter sequence, and the whole network average power parameter includes a whole network average active power and a whole network average reactive power; A difference value between a local active power and a whole network average active power of each distributed unit at each time point is sequentially calculated to obtain an active power deviation value at each time point, and a difference value between a local reactive power and a whole network average reactive power of each distributed unit at each time point is sequentially calculated to obtain a reactive power deviation value at each time point; Integral operations are respectively performed on the active power deviation values and the reactive power deviation values of all time points of each distributed unit to obtain a cumulative active power deviation amount and a cumulative reactive power deviation amount of each distributed unit in the time period, and the cumulative active power deviation amount and the cumulative reactive power deviation amount are output as the power deviation values of each distributed unit.

6. The time synchronization control method for a distributed power generation system according to claim 1, characterized by, According to the power deviation values, a regulation signal of each distributed unit is generated, and a distributed power generation system is regulated and optimized according to the regulation signal, specifically as follows: Based on the power deviation values, an active power deviation sequence and a reactive power deviation sequence of each distributed unit are constructed, a fuzzy PID control algorithm is introduced, and initial values of proportional coefficients, integral coefficients and differential coefficients of the fuzzy PID control algorithm are set; Active power deviation change rates and reactive power deviation change rates are respectively calculated according to the active power deviation sequence and the reactive power deviation sequence; The active power deviation values and the reactive power deviation values are taken as input variables of the fuzzy PID control algorithm, the proportional coefficients, the integral coefficients and the differential coefficients are adaptively adjusted through a fuzzy rule base, and optimized proportional coefficients, integral coefficients and differential coefficients are obtained. According to the optimized proportional coefficient, integral coefficient and differential coefficient, active power regulation signals and reactive power regulation signals of each distributed unit are calculated, the active power regulation signal is a speed regulation control signal, and the reactive power regulation signal is a voltage regulation control signal; The speed regulation control signal is output to a speed regulator of each distributed unit to change the active power output of the generator by adjusting the fuel supply amount or the steam admission amount of the prime mover, and the voltage regulation control signal is output to an automatic voltage regulator of each distributed unit to change the terminal voltage and the reactive power output of the generator by adjusting the field current; Real-time monitoring of the active power and reactive power parameters of each distributed unit after regulation, recalculation of the power deviation value, and if the power deviation value does not reach the preset convergence threshold, the fuzzy PID regulation process is repeated until the power deviation value of each distributed unit is less than the convergence threshold.

7. A time synchronization control system for a distributed power generation system, characterized by, The time synchronization control system for the distributed power generation system comprises a storage and a processor, the storage comprises a time synchronization control method program for the distributed power generation system, and the time synchronization control method program for the distributed power generation system is executed by the processor to realize the following steps: Local clock deviation data of the generators of each distributed unit in the distributed power generation system is acquired, each distributed unit is time-corrected according to the local clock deviation data, the running power data of each distributed unit after time correction is collected in real time, and a local power parameter sequence with a time label is generated; The local power parameter sequence is packaged into a synchronization message, the synchronization message is transmitted to the master control end, and the synchronization message is pre-distributed to obtain response data of each distributed unit to the pre-distributed synchronization message; According to the response data, the receiving integrity of each distributed unit to the synchronization message at the same time point is determined, and the distribution strategy of the synchronization message is determined according to the integrity; According to the distribution strategy, the synchronization message is distributed to each distributed unit, and the power deviation value of the local power parameter and the network average power parameter is calculated; According to the power deviation value, a regulation signal of each distributed unit is generated, and the distributed power generation system is regulated and optimized according to the regulation signal.

Images