A method and system for intelligent control of parallel operation of generator sets
By obtaining the rated reactive power and excitation current of the generator sets, a benchmark unit is selected and refined reactive power adjustment is performed, which solves the problem of reactive power circulation between heterogeneous units and improves the stability and response performance of the generator sets in parallel operation.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- BEIJING SHENGBOTE MECHANICAL & ELECTRICAL EQUIP CO LTD
- Filing Date
- 2026-04-24
- Publication Date
- 2026-07-21
Smart Images
Figure CN122437126A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of generator control technology, specifically to an intelligent control method and system for parallel operation of generator sets. Background Technology
[0002] With the increasing demands for stability, automation, and intelligence in modern power systems, generator parallel operation control technology is constantly developing and improving, especially as the demand for electricity grows during industrialization and urbanization. At the same time, the scale and complexity of power systems are also constantly increasing. Factors such as demand fluctuations, equipment aging, and load changes during power production and transmission pose challenges. To address these challenges, modern power systems have gradually evolved into a mode of parallel operation of multiple generator units, with multiple units working simultaneously to improve generation capacity and ensure the reliability of power supply.
[0003] In current parallel operation control of multiple generator units, a thorny engineering challenge is the reactive power circulating current problem between heterogeneous units. When generators of different brands, capacities, or depreciation levels are connected in parallel, the differences in voltage regulation characteristics and phase deviations of their respective excitation systems lead to continuous reactive power circulating currents between the units. This circulating current not only increases the heat loss of the stator windings and reduces power generation efficiency, but in severe cases, it can also trigger the reverse power protection of the motor, causing unplanned shutdowns. In traditional droop control mode, due to the lack of a precise reactive power adjustment mechanism, the circulating current is often difficult to eliminate autonomously, resulting in unstable control effects. Summary of the Invention
[0004] To address the issue of unstable control performance in existing methods for intelligent regulation of generator parallel operation, the present invention aims to provide an intelligent control method and system for generator parallel operation. The specific technical solution adopted is as follows: In a first aspect, the present invention provides a method for intelligent control of generator sets in parallel operation, the method comprising the following steps: Obtain the rated reactive power and rated excitation current of each generator set, and determine the reactive power trend value; Based on the rated reactive power, rated excitation current, and cumulative operating time of each key load point in each generator set, the relative deviation score of the excitation characteristics of each generator set is obtained; the operating time score is determined based on the operating time of each generator set; and the benchmark generator set is selected by combining the relative deviation score of the excitation characteristics and the operating time score. Based on the deviation of the reactive power trend value between each generator set and the benchmark generator set, and combined with the fluctuation characteristics of the reactive power trend value of each generator set, it is determined whether to start the adjustment process. When the regulation process is started, the corresponding PI control parameters are called according to the load rate range to calculate the theoretical correction amount of reactive power deviation. After limiting the theoretical correction amount, the excitation current setpoint is output to the automatic voltage regulator of the corresponding generator set.
[0005] Preferably, obtaining the reactive power trend value includes: Calculate the reactive power for each sampling period based on the instantaneous terminal voltage and instantaneous current values of each generator set in each sampling period. The reactive power is averaged using a sliding window for a first number of consecutive sampling periods to obtain the steady-state trend value of the reactive power.
[0006] Preferably, the step of obtaining the relative deviation score of the excitation characteristics of each generator set based on the rated reactive power, rated excitation current, and cumulative operating time at each key load point in each generator set includes: Obtain the deviation between the rated reactive power and the steady-state value of reactive power corresponding to the rated excitation current of the candidate generator set at each critical load point; The ratio of the cumulative operating time within the load range corresponding to each key load point of the candidate generator set to the total cumulative operating time is determined as the weight of each key load point; the excitation characteristic deviation is obtained based on the deviation value and the weight. Based on the deviation of the excitation characteristics of the candidate generator sets, the relative deviation score of the excitation characteristics of the candidate generator sets is obtained. The candidate generator set can be any generator set.
[0007] Preferably, the determination of runtime score based on the runtime of each generator set includes: A runtime score is calculated based on the ratio of the actual runtime of the candidate generator set to the preset maximum runtime, and the runtime and the runtime score are negatively correlated.
[0008] Preferably, obtaining the excitation characteristic deviation based on the deviation value and the weight includes: The weighted sum of each deviation value and its corresponding weight is taken as the deviation of the excitation characteristic.
[0009] Preferably, the comprehensive excitation characteristic relative deviation score and runtime score are used to screen benchmark units, including: For any generator set: Multiply the excitation characteristic relative deviation score and the running time score of the generator set by the corresponding weighting coefficients and then sum them to obtain the comprehensive operating status score of the generator set. The generator set with the highest comprehensive operating status score is used as the benchmark unit.
[0010] Preferably, the step of determining whether to initiate the adjustment process based on the deviation of the reactive power trend value of each generator set from that of the benchmark generator set, combined with the fluctuation characteristics of the reactive power trend value of each generator set, includes: For any given generator set: Calculate the reactive power trend value deviation between any generator set and the benchmark generator set in the same sampling period; Based on the standard deviation of the reactive power deviation of any generator set within a consecutive preset second number of sampling periods, determine the upper limit of the allowable reactive power deviation and the hysteresis during steady-state operation. When the absolute value of the reactive power deviation between any generator set and the reference generator set in each sampling period is greater than the hysteresis, the adjustment process is initiated. When the absolute value of the reactive power deviation between any generator set and the reference generator set in each sampling period is less than or equal to the upper limit of the reactive power deviation allowed during steady-state operation, the adjustment process is stopped. When the absolute value of the reactive power deviation between any generator set and the reference generator set in each sampling period is between the upper limit of the reactive power deviation allowed during steady-state operation and the hysteresis, the adjustment state of the previous period is maintained.
[0011] Preferably, the step of calling the corresponding PI control parameters according to the load rate range to calculate the theoretical correction amount of reactive power deviation includes: For any given generator set: The load rate is divided into multiple intervals, and each interval is configured with a proportional coefficient and an integral coefficient. The interval to which any generator set belongs is determined based on its load rate, and the corresponding proportional coefficient and integral coefficient are called. An incremental PI algorithm is used to calculate the theoretical correction amount based on the reactive power deviation of the current sampling period and the reactive power deviation of the previous sampling period.
[0012] Preferably, the method of limiting the theoretical correction amount and then outputting the excitation current setpoint to the automatic voltage regulator of the corresponding generator set includes: Obtain the minimum value between the absolute value of the theoretical correction and the preset maximum allowable step size; Based on the theoretical correction and the minimum value, the actual execution step size of any generator set is obtained, and the direction of increase or decrease of the excitation current setpoint is determined according to the direction of reactive power deviation. The actual execution step size is then superimposed on the excitation current setpoint of the previous sampling period and output to the automatic voltage regulator.
[0013] Secondly, the present invention provides an intelligent control system for parallel operation of generator sets, the system comprising: The data acquisition module is used to acquire the rated reactive power and rated excitation current of each generator set, and to determine the reactive power trend value. The benchmark screening module is used to obtain the relative deviation score of the excitation characteristics of each generator set based on the rated reactive power, rated excitation current and cumulative operating time of each key load point in each generator set; determine the operating time score based on the operating time of each generator set; and screen benchmark units by combining the relative deviation score of excitation characteristics and the operating time score. The judgment module is used to determine whether to start the adjustment process based on the deviation of the reactive power trend value of each generator set from that of the benchmark generator set, combined with the fluctuation characteristics of the reactive power trend value of each generator set. The control module is used to call the corresponding PI control parameters according to the load rate range when the control process is started, calculate the theoretical correction amount of reactive power deviation, limit the theoretical correction amount, and output the excitation current setpoint to the automatic voltage regulator of the corresponding generator set.
[0014] The present invention has at least the following beneficial effects: This invention accurately reflects the real-time reactive power output status of each generator set by acquiring the rated reactive power and rated excitation current of each generator set and determining the reactive power trend value. A benchmark generator set is selected by comprehensively considering the relative deviation score of excitation characteristics and the running time score, ensuring that the selection of the benchmark generator set is based on sufficient and dynamic rationality, thus improving the reference stability of the control system. Furthermore, the deviation and fluctuation characteristics of the reactive power trend value between each generator set and the benchmark generator set determine whether to initiate the adjustment process, avoiding frequent adjustments caused by instantaneous disturbances or noise interference, and enhancing the anti-interference capability and judgment accuracy of the control process. When the adjustment process is initiated, the corresponding PI control parameters are called according to the load rate range, the theoretical correction amount of the reactive power deviation is calculated, and after limiting it, the excitation current setpoint is output to the automatic voltage regulator. This achieves refined and interval-based adjustment of reactive power, suppresses reactive power circulating current in the parallel system, and improves the stability and dynamic response performance of the generator set operation control. Attached Figure Description
[0015] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0016] Figure 1 A flowchart of a generator set parallel intelligent control method provided in an embodiment of the present invention; Figure 2 This is a structural block diagram of a generator set parallel intelligent control system provided in an embodiment of the present invention. Detailed Implementation
[0017] To further illustrate the technical means and effects adopted by the present invention to achieve the intended purpose, the following detailed description of a generator set parallel intelligent control method and system proposed according to the present invention is provided in conjunction with the accompanying drawings and preferred embodiments.
[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.
[0019] The following description, in conjunction with the accompanying drawings, details the specific scheme of the intelligent control method and system for parallel operation of generator sets provided by the present invention.
[0020] An embodiment of a generator set parallel intelligent control method: This embodiment proposes a method for intelligent control of generator sets in parallel operation, such as... Figure 1 As shown, a generator set parallel intelligent control method according to this embodiment includes the following steps: Step S1: Obtain the rated reactive power and rated excitation current of each generator set, and determine the reactive power trend value.
[0021] During the operation of generator sets in an islanded microgrid or a small-scale grid-connected system of equivalent capacity with a master-slave control architecture, the instantaneous terminal voltage and current values of each generator set are synchronously collected through voltage transformers and current transformers. In this embodiment, a sampling period of 50ms is used, the sampling rate is 6.4kHz, and each sampling period contains 128 sampling time points. In specific applications, implementers can set the parameters according to specific circumstances.
[0022] The correlation between the excitation current setpoint and the steady-state trend value of reactive power for each generator set under no-load, 25%, 50%, 75%, and 100% rated load conditions is obtained. An excitation characteristic curve for each generator set is then fitted and stored in a historical database as a characteristic fingerprint. Simultaneously, during normal operation, the excitation current setpoint and the steady-state trend value of reactive power under the current stable operating condition are added to the database every 10 minutes, and the curve is refitted and updated to ensure that the fingerprint reflects characteristic drift caused by factors such as aging.
[0023] Obtain the excitation characteristic curve provided by the manufacturer when the generator set leaves the factory, and use it to obtain the rated excitation current and rated reactive power at several key load points (such as no-load, 25%, 50%, 75%, and 100% load).
[0024] Based on the instantaneous terminal voltage and current values of each generator set in each sampling period, the reactive power for each sampling period is calculated. The calculation method for reactive power is existing technology and will not be elaborated further in this embodiment. A sliding window averaging process is applied to the reactive power over a preset first number of sampling periods to remove abnormal peaks caused by instantaneous load fluctuations (such as large motor startup), and the steady-state trend value of reactive power is extracted. In this embodiment, the preset first number is 5; in specific applications, the implementer can set it according to specific circumstances.
[0025] Step S2: Based on the rated reactive power, rated excitation current and cumulative operating time of each key load point in each generator set, obtain the relative deviation score of the excitation characteristics of each generator set; determine the operating time score based on the operating time of each generator set; and select the benchmark generator set by combining the relative deviation score of the excitation characteristics and the operating time score.
[0026] When performing parallel control of generator sets, the traditional method fixes one generator set as a reference. If the performance of the generator set itself deteriorates, such as the excitation system aging, its reactive power output may deviate from the normal value. Using it as a reference will affect the accuracy of the subsequent analysis results. Therefore, this embodiment will ensure that the reference is always a relatively stable generator set that meets the ideal characteristics through dynamic updates.
[0027] This embodiment considers two characteristics for selecting a relatively ideal unit: its characteristics are close to the factory value (representing a healthy excitation system, and the relationship between reactive power output and excitation current is in line with expectations, making it a highly reliable benchmark) and its short operating time (new units have less aging and less parameter drift). The combination of these two characteristics ensures the relative physical stability of the benchmark unit.
[0028] For each critical load point in the generator set, the closer it is to the commonly used load range, the more attention should be paid to whether the excitation current relationship corresponding to that critical load point is more in line with expectations.
[0029] The following embodiment uses a generator set as an example for explanation. The method provided in this embodiment can be used to process other generator sets.
[0030] Specifically, any generator set is designated as a candidate generator set.
[0031] Based on the excitation current curves of candidate generator sets, the rated reactive power corresponding to the rated excitation current at each critical load point is obtained. The absolute value of the difference between the rated reactive power corresponding to the rated excitation current at each critical load point is taken as the deviation value between the rated reactive power corresponding to the rated excitation current at each critical load point and the steady-state value of reactive power. The ratio of the cumulative operating time within the load range corresponding to each critical load point of the candidate generator set to the total cumulative operating time is determined as the weight of each critical load point. The weighted sum of each deviation value and its corresponding weight is taken as the excitation characteristic deviation.
[0032] As a specific example, the formula for calculating the deviation of excitation characteristics is given. The formula for calculating the deviation of excitation characteristics of candidate generator sets is as follows: in, This indicates that the excitation characteristics of the candidate generator set deviate relatively from the score. This indicates the number of critical load points for the candidate generator set. This represents the deviation between the rated reactive power and the steady-state value of reactive power corresponding to the rated excitation current at the nth critical load point of the candidate generator set. This represents the weight of the nth critical load point of the candidate generator set. This represents the normalization function. In this embodiment, the maximum-minimum normalization method is used to normalize the data. In specific applications, other existing data normalization methods can also be used.
[0033] Furthermore, based on the excitation characteristic deviation of the candidate generator sets, a relative deviation score for the excitation characteristics of the candidate generator sets is obtained. Specifically, the maximum value of the excitation characteristic deviation of all generator sets is obtained, and the ratio between the excitation characteristic deviation of the candidate generator set and this maximum value is calculated. The absolute value of the difference between the constant 1 and this ratio is taken as the relative deviation score of the candidate generator set's excitation characteristics. It should be noted that when calculating the relative deviation score of the excitation characteristics, if the denominator is 0, a very small positive number, such as 0.0001, is added to the denominator before recalculation.
[0034] The runtime score is calculated based on the ratio of the actual runtime of the candidate generator set to the preset maximum runtime. The runtime and the runtime score are negatively correlated.
[0035] As a concrete example, the specific formula for calculating runtime score is given. Runtime score can be expressed as: in, Indicates runtime score. This indicates the actual operating time of the candidate generator set. Indicates the preset maximum runtime. This represents the function that takes the maximum value.
[0036] In this embodiment, the maximum runtime is preset to 50,000 hours. In specific applications, the implementer can set this value according to the specific circumstances. It should be noted that when calculating the runtime score, if the denominator is 0, a very small positive number is added to the denominator before recalculation, such as 0.0001, whose dimension is the same as that of the runtime.
[0037] For candidate generator sets: The comprehensive operating status score of the generator set is obtained by multiplying the excitation characteristic relative deviation score and the running time score of the generator set by the corresponding weighting coefficients.
[0038] As a specific example, the formula for calculating the comprehensive operational status score is given. The comprehensive operational status score can be expressed as: in, This represents the overall operating status score of the candidate generator units. This represents the weighting coefficient indicating the relative deviation of the excitation characteristics from the score. This represents the weighting coefficient for runtime scoring.
[0039] In this embodiment, the weighting coefficient for the relative deviation score of excitation characteristics is 0.6, and the weighting coefficient for the runtime score is 0.4. In specific applications, implementers can set these values according to specific circumstances.
[0040] Using the above methods, a comprehensive operating status score for each generator set can be obtained, and the generator set with the highest comprehensive operating status score is selected as the benchmark unit. In this embodiment, the comprehensive operating status score of all generator sets is recalculated every 15 minutes, and the benchmark unit is re-determined, achieving dynamic updates to the benchmark and ensuring that the benchmark unit is the most stable operating unit.
[0041] Step S3: Based on the deviation of the reactive power trend value between each generator set and the benchmark generator set, and combined with the fluctuation characteristics of the reactive power trend value of each generator set, determine whether to start the adjustment process.
[0042] In this embodiment, a reference generator set is determined in step S2. Next, the deviation between the reactive power trend values of each generator set and the reference generator set will be compared, and the fluctuation characteristics of the reactive power trend values of each generator set will be combined to determine whether the adjustment process needs to be initiated.
[0043] The following explanation will use a single generator set as an example. The method provided in this embodiment can be used to process other generator sets as well.
[0044] Specifically, for any given generator set: For any sampling period, the reactive power trend value of the generator set in that sampling period is calculated, and the reactive power deviation of the reactive power trend value of the reference unit in that sampling period is subtracted from the reactive power trend value of the reference unit in that sampling period. This difference is taken as the reactive power deviation between the generator set and the reference unit in that sampling period. When the reactive power deviation is positive, it indicates that the reactive power output of the generator set is too high; when the reactive power deviation is negative, it indicates that the reactive power output of the generator set is too low.
[0045] Next, the dead zone needs to be determined to prevent frequent adjustments caused by minor fluctuations and extend the AVR's lifespan. However, if the dead zone is fixed, it may be too conservative or sensitive when noise changes. Therefore, real-time noise statistics can be used to calculate the dead zone so that it always matches the current environment. At the same time, the hysteresis (greater than the dead zone) is determined to create a high start-up threshold and a low stop-down threshold, avoiding repeated start-stop cycles caused by deviations crossing the dead zone boundary. Adaptive hysteresis is used to improve anti-interference capability.
[0046] Considering that the standard deviation of the generator set will be extremely large during a normal load change transient process, leading to an excessively large dead zone and control failure, the dead zone and hysteresis are updated when the active power change rate of each generator set is lower than a set threshold. Simultaneously, a dead zone upper limit is set. If the calculated upper limit of the allowable reactive power deviation during steady-state operation exceeds the set dead zone upper limit, the set dead zone upper limit is used. In this embodiment, the dead zone upper limit is set to 2% of the generator set's rated reactive power. In specific applications, the implementer can set this value according to specific circumstances.
[0047] Specifically, based on the standard deviation of the reactive power deviation of the generator set within a consecutive preset second number of sampling periods, the upper limit of the allowable reactive power deviation and the hysteresis error during steady-state operation are determined. Specifically, twice the standard deviation is used as the upper limit of the allowable reactive power deviation during steady-state operation, and three times the standard deviation is used as the hysteresis error. In this embodiment, the preset second number is 20; in specific applications, the implementer can set it according to specific circumstances. It should be noted that: in this embodiment, when analyzing a sampling period, the consecutive preset second number of sampling periods refers to the consecutive preset second number of sampling periods before and adjacent to the current sampling period. The upper limit of the allowable reactive power deviation during steady-state operation is set at 2 times the standard deviation, which is determined based on the statistical characteristics of the normal distribution. This range covers more than 95% of the measurement noise. Deviation fluctuations within this range are highly likely to be normal noise rather than real circulating current. Therefore, setting it as a dead zone can avoid frequent malfunctions of the AVR. The hysteresis is set at 3 times the standard deviation, which corresponds to a 99.7% confidence interval. As a start-up threshold, it can ensure that readjustment is only performed when the deviation significantly exceeds the noise background. At the same time, it forms a hysteresis band with the dead zone corresponding to 2 times the standard deviation, preventing the deviation from crossing back and forth at the dead zone boundary and causing repeated start-stop cycles.
[0048] When the absolute value of the reactive power deviation between the generator set and the reference unit in the sampling period is greater than the hysteresis, the adjustment process is initiated; when the absolute value of the reactive power deviation between the generator set and the reference unit in the sampling period is less than or equal to the upper limit of the reactive power deviation allowed during steady-state operation, the adjustment process is stopped; when the absolute value of the reactive power deviation between the generator set and the reference unit in the sampling period is between the upper limit of the reactive power deviation allowed during steady-state operation and the hysteresis, the adjustment state of the previous period is maintained.
[0049] Step S4: When the adjustment process is started, the corresponding PI control parameters are called according to the load rate range to calculate the theoretical correction amount of reactive power deviation, and the excitation current setpoint is output to the automatic voltage regulator of the corresponding generator set after the theoretical correction amount is limited.
[0050] Considering the different dynamic characteristics of generators under different load rates—low damping under light loads, excessive proportional gain can easily cause oscillations; and high inertia under heavy loads, requiring a stronger proportional gain for rapid response—parameters are optimized according to load rate zones, and corrected through feedback from actual adjustment effects, allowing parameters to automatically adapt to various operating conditions. Simultaneously, the entire process is continuous, constantly tracking system changes.
[0051] First, based on each generator unit, the load rate for each sampling period can be determined according to the active power and the rated power of the unit.
[0052] The following explanation will use a single generator set as an example. The method provided in this embodiment can be used to process other generator sets as well.
[0053] Specifically, for any given generator set: The load rate of 0-100% is evenly divided into multiple intervals. In this embodiment, the number of intervals is 10. In specific applications, the implementer can set the interval according to the specific situation.
[0054] This embodiment uses a PI controller for control. The reason for using a PI controller instead of a PID controller is that the excitation system has a large inertia and the reactive power regulation is sensitive to noise. The derivative term can easily amplify measurement noise, leading to system instability. PI control is sufficient to achieve zero steady-state error tracking and has better robustness.
[0055] For each load range, a set of corresponding PI parameters to be optimized are configured, namely the proportional coefficient and the integral coefficient. The PI parameters in all load ranges are initially set to the same set of empirical values. In this embodiment, the initial value of the proportional coefficient for each range is set to 1.0, and the initial value of the integral coefficient is set to 2.0. In specific applications, these values can be set according to the generator set capacity. These initial values are only used for startup and will be quickly optimized through online learning.
[0056] Each time a complete adjustment process ends (i.e., from the deviation exceeding the threshold to entering the dead zone and stabilizing), the average load rate, maximum overshoot, and adjustment time of that adjustment are recorded. The overshoot is defined as the ratio of the peak deviation to the initial deviation at the start of the adjustment; the adjustment time is defined as the time from the start of the adjustment (when the reactive power deviation of the generator unit relative to the benchmark unit first exceeds the abnormal circulating current monitoring threshold of the generator unit) to the first time the deviation enters the dead zone. Simultaneously, the load range to which the adjustment belongs is determined based on the average load rate of that adjustment, and the parameters of that range are corrected according to the following rules: If the overshoot is greater than the target overshoot, it indicates that the proportional gain is too strong, so the proportional coefficient is appropriately reduced, i.e.: That is, reducing the proportional coefficient of that interval to its original value. times, of which, This is the proportionality coefficient. To adjust the step size, in this embodiment, the target overshoot is set to 10%. If the settling time is greater than the target settling time, it indicates that the response is too slow, so the proportional coefficient is appropriately increased. That is, increasing the proportional coefficient of that interval to the original value. The target adjustment time is 2 seconds; in this embodiment, the target adjustment time is 2 times.
[0057] The integral coefficient is adjusted according to its proportional relationship with the proportional coefficient, keeping the ratio between the integral coefficient and the proportional coefficient essentially constant. In this embodiment, this constant value is 0.2. In specific applications, the implementer can set it according to the specific situation. The correction step size is generally between 0.01 and 0.1. In this embodiment, the initial preset value is 0.05. This value, as the step size for online learning, ensures both the sensitivity of parameter correction and, in conjunction with the attenuation formula (e.g., attenuating by 10% with each adjustment), ensures that the PI parameter eventually converges to a stable value, avoiding continuous oscillations in the control system.
[0058] To prevent the parameters from decaying indefinitely to zero or diverging indefinitely under extreme operating conditions where the set overshoot target cannot be reached, a verification is performed after each parameter correction. and Does it exceed the preset allowable range? If the value exceeds the limit, it is truncated to the boundary value, and the corresponding boundary value is used as the corrected parameter. This represents the lower limit of the preset allowed range. This represents the upper limit of the preset allowed range. In this embodiment, it is set... It is 0.1. The value is set to 5.0. This range setting can prevent the PI parameter from failing due to excessive decay, and can also prevent the system from resonating or overshooting due to excessively large parameters, ensuring that the online learning process is always within a safe and stable gain range.
[0059] For the i-th generator set, based on the PI parameters obtained from the above process, the incremental PI algorithm is used to calculate the theoretical correction amount according to the reactive power deviation of the current sampling period and the reactive power deviation of the previous sampling period. Only the change relative to the previous period is output, instead of the absolute value, thereby avoiding integral saturation and facilitating amplitude limiting.
[0060] As a concrete example, the specific formula for calculating the theoretical correction is given. The theoretical correction for the i-th generator unit can be expressed as: in, This represents the theoretical correction amount for the i-th generator set; This represents the reactive power deviation of the i-th generator set relative to the reference generator set during the current cycle. This represents the reactive power deviation of the i-th generator set relative to the reference generator set in the previous cycle of the current cycle. This indicates the proportional coefficient for the load range to which this adjustment belongs. This indicates the integral coefficient for the load range to which this adjustment belongs.
[0061] The theoretical correction is used to characterize the extent to which the excitation current setpoint should be adjusted to reduce reactive power deviation, and is expressed as a percentage of rated excitation. A positive theoretical correction indicates that excitation needs to be increased, while a negative theoretical correction indicates that excitation needs to be decreased.
[0062] To prevent oscillations caused by excessively large single adjustment amplitudes, the theoretical correction amount calculated above needs to be limited.
[0063] Specifically, the minimum value between the absolute value of the theoretical correction and the preset maximum allowable step size is obtained. In this embodiment, the preset maximum allowable step size is set to 0.01, meaning that the maximum allowable change in excitation current during a single adjustment does not exceed 1% of the rated value. This setting serves as a hard limit for the incremental PI output, effectively preventing system oscillations caused by excessively rapid single adjustments and ensuring the physical safety of the parallel system during dynamic adjustment. In practical applications, the implementer can set the preset maximum allowable step size according to specific circumstances. Then, based on the theoretical correction and the minimum value between the theoretical correction and the preset maximum allowable step size, the actual execution step size of the generator set is obtained. The direction of increase or decrease of the excitation current setpoint is determined according to the direction of reactive power deviation. The actual execution step size is then superimposed on the excitation current setpoint of the previous sampling period and output to the automatic voltage regulator.
[0064] As a concrete example, the specific formula for calculating the actual execution step size is given. The current actual execution step size of the i-th generator set can be expressed as: in, This represents the actual execution step size of the i-th generator set in the current cycle. This represents the theoretical correction amount for the i-th generator set. This indicates the preset maximum allowable step size. This represents the function that takes the minimum value. Indicates the absolute value sign. Represents a symbolic function.
[0065] Incremental PI converters only output the change to avoid integral saturation, while the correction is executed after step-size limiting to ensure safety. In this embodiment, the actual execution step size is superimposed on the current excitation setpoint based on the deviation direction. Specifically: if the reactive power deviation between the i-th generator set and the reference generator set in the current sampling period is greater than 0, it indicates high reactive power; if the reactive power deviation between the i-th generator set and the reference generator set in the current sampling period is less than 0, it indicates low reactive power. Based on the above characteristics, let: ,in, This represents the excitation current setpoint sent to the i-th unit AVR during the current sampling period; This represents the excitation current setpoint sent to the i-th generator AVR in the previous sampling period of the current sampling period; This represents the actual execution step size of the current cycle for the i-th generator set.
[0066] The AVR controls the generator's magnetic field strength by actually outputting the corresponding excitation current through the excitation system based on the given excitation current value. Simultaneously, after adjustment, it waits two sampling cycles (100ms) before proceeding to the next comparison to allow the excitation system to respond and stabilize. The 100ms wait is because the excitation system is an electromechanical transient process; changing the given value requires time to respond. Sampling during this period would yield transient values, leading to erroneous judgments. Therefore, waiting two cycles allows the system to enter a new steady state, ensuring the accuracy of the next sampling.
[0067] Thus, by using the method provided in this embodiment, intelligent control of the generator set has been achieved.
[0068] This embodiment obtains the rated reactive power and rated excitation current of each generator set and determines the reactive power trend value, which can accurately reflect the real-time reactive power output status of each unit. A benchmark generator set is selected by comprehensively considering the relative deviation score of excitation characteristics and the running time score, ensuring that the selection of the benchmark generator set is based on sufficient and dynamic rationality, thus improving the reference stability of the parallel control system. Furthermore, the deviation and fluctuation characteristics of the reactive power trend value between each generator set and the benchmark generator set determine whether to initiate the adjustment process, avoiding frequent adjustments caused by instantaneous disturbances or noise interference, and enhancing the anti-interference capability and judgment accuracy of the control process. When the adjustment process is initiated, the corresponding PI control parameters are called according to the load rate range, the theoretical correction amount of the reactive power deviation is calculated, and after limiting it, the excitation current setpoint is output to the automatic voltage regulator. This achieves refined and interval-based adjustment of reactive power, suppresses reactive power circulating current in the parallel system, and improves the control stability and dynamic response performance of the generator set operation.
[0069] An embodiment of a generator set parallel intelligent control system: See Figure 2 The diagram illustrates a structural block diagram of a generator set parallel intelligent control system according to an embodiment of the present invention. The system may include a data acquisition module, a benchmark screening module, a judgment module, and a control module.
[0070] The data acquisition module is used to acquire the rated reactive power and rated excitation current of each generator set, and to determine the reactive power trend value. The benchmark screening module is used to obtain the relative deviation score of the excitation characteristics of each generator set based on the rated reactive power, rated excitation current and cumulative operating time of each key load point in each generator set; determine the operating time score based on the operating time of each generator set; and screen benchmark units by combining the relative deviation score of excitation characteristics and the operating time score. The judgment module is used to determine whether to start the adjustment process based on the deviation of the reactive power trend value of each generator set from that of the benchmark generator set, combined with the fluctuation characteristics of the reactive power trend value of each generator set. The control module is used to call the corresponding PI control parameters according to the load rate range when the control process is started, calculate the theoretical correction amount of reactive power deviation, limit the theoretical correction amount, and output the excitation current setpoint to the automatic voltage regulator of the corresponding generator set.
[0071] It should be understood that Figure 2 The structural block diagram and modules of the generator parallel intelligent control system shown can be implemented in various ways. For example, in some embodiments, the system and its modules can be implemented by hardware, software, or a combination of software and hardware. The hardware portion can be implemented using dedicated logic; the software portion can be stored in memory and executed by appropriate instructions, such as a microprocessor or dedicated hardware. Those skilled in the art will understand that the above-described methods and systems can be implemented using computer-executable instructions and / or included in processor control code, for example, on a carrier medium such as a disk, CD, or DVD-ROM, a programmable memory such as read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The systems and modules of this specification can be implemented not only by hardware circuits such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field-programmable gate arrays, programmable logic devices, etc., but also by software executed by various types of processors, or by a combination of the above-described hardware circuits and software (e.g., firmware).
[0072] For more details about the above modules, please refer to other parts of this manual; they will not be repeated here.
[0073] It should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A method for intelligent control of generator sets in parallel operation, characterized in that, The method includes the following steps: Obtain the rated reactive power and rated excitation current of each generator set, and determine the reactive power trend value; Based on the rated reactive power, rated excitation current, and cumulative operating time of each key load point in each generator set, the relative deviation score of the excitation characteristics of each generator set is obtained; the operating time score is determined based on the operating time of each generator set; and the benchmark generator set is selected by combining the relative deviation score of the excitation characteristics and the operating time score. Based on the deviation of the reactive power trend value between each generator set and the benchmark generator set, and combined with the fluctuation characteristics of the reactive power trend value of each generator set, it is determined whether to start the adjustment process. When the regulation process is started, the corresponding PI control parameters are called according to the load rate range to calculate the theoretical correction amount of reactive power deviation. After limiting the theoretical correction amount, the excitation current setpoint is output to the automatic voltage regulator of the corresponding generator set.
2. The intelligent control method for parallel operation of generator sets according to claim 1, characterized in that, The acquisition of the reactive power trend value includes: Calculate the reactive power for each sampling period based on the instantaneous terminal voltage and instantaneous current values of each generator set in each sampling period. The reactive power is averaged using a sliding window for a first number of consecutive sampling periods to obtain the steady-state trend value of the reactive power.
3. The intelligent control method for parallel operation of generator sets according to claim 1, characterized in that, The relative deviation score of the excitation characteristics of each generator set is obtained based on the rated reactive power, rated excitation current, and cumulative operating time at each key load point in each generator set, including: Obtain the deviation between the rated reactive power and the steady-state value of reactive power corresponding to the rated excitation current of the candidate generator set at each critical load point; The ratio of the cumulative operating time within the load range corresponding to each key load point of the candidate generator set to the total cumulative operating time is determined as the weight of each key load point; the excitation characteristic deviation is obtained based on the deviation value and the weight. Based on the deviation of the excitation characteristics of the candidate generator sets, the relative deviation score of the excitation characteristics of the candidate generator sets is obtained. The candidate generator set can be any generator set.
4. The intelligent control method for parallel operation of generator sets according to claim 3, characterized in that, The determination of runtime score based on the runtime of each generator set includes: A runtime score is calculated based on the ratio of the actual runtime of the candidate generator set to the preset maximum runtime, and the runtime and the runtime score are negatively correlated.
5. The intelligent control method for parallel operation of generator sets according to claim 3, characterized in that, The step of obtaining the excitation characteristic deviation based on the deviation value and the weight includes: The weighted sum of each deviation value and its corresponding weight is taken as the deviation of the excitation characteristic.
6. The intelligent control method for parallel operation of generator sets according to claim 1, characterized in that, The comprehensive excitation characteristic relative deviation score and operating time score are used to screen benchmark units, including: For any generator set: Multiply the excitation characteristic relative deviation score and the running time score of the generator set by the corresponding weighting coefficients and then sum them to obtain the comprehensive operating status score of the generator set. The generator set with the highest comprehensive operating status score is used as the benchmark unit.
7. The intelligent control method for parallel operation of generator sets according to claim 1, characterized in that, The determination of whether to initiate the adjustment process based on the deviation of the reactive power trend value of each generator set from that of the benchmark generator set, combined with the fluctuation characteristics of the reactive power trend value of each generator set, includes: For any given generator set: Calculate the reactive power trend value deviation between any generator set and the benchmark generator set in the same sampling period; Based on the standard deviation of the reactive power deviation of any generator set within a consecutive preset second number of sampling periods, determine the upper limit of the allowable reactive power deviation and the hysteresis during steady-state operation. When the absolute value of the reactive power deviation between any generator set and the reference generator set in each sampling period is greater than the hysteresis, the adjustment process is initiated. When the absolute value of the reactive power deviation between any generator set and the reference generator set in each sampling period is less than or equal to the upper limit of the reactive power deviation allowed during steady-state operation, the adjustment process is stopped. When the absolute value of the reactive power deviation between any generator set and the reference generator set in each sampling period is between the upper limit of the reactive power deviation allowed during steady-state operation and the hysteresis, the adjustment state of the previous period is maintained.
8. The intelligent control method for parallel operation of generator sets according to claim 1, characterized in that, The step of calculating the theoretical correction amount for reactive power deviation by calling the corresponding PI control parameters according to the load rate range includes: For any given generator set: The load rate is divided into multiple intervals, and each interval is configured with a proportional coefficient and an integral coefficient. The interval to which any generator set belongs is determined based on its load rate, and the corresponding proportional coefficient and integral coefficient are called. An incremental PI algorithm is used to calculate the theoretical correction amount based on the reactive power deviation of the current sampling period and the reactive power deviation of the previous sampling period.
9. The intelligent control method for parallel operation of generator sets according to claim 1, characterized in that, After limiting the theoretical correction, the excitation current setpoint is output to the automatic voltage regulator of the corresponding generator set, including: Obtain the minimum value between the absolute value of the theoretical correction and the preset maximum allowable step size; Based on the theoretical correction and the minimum value, the actual execution step size of any generator set is obtained, and the direction of increase or decrease of the excitation current setpoint is determined according to the direction of reactive power deviation. The actual execution step size is then superimposed on the excitation current setpoint of the previous sampling period and output to the automatic voltage regulator.
10. A generator set paralleling intelligent control system, the system being used to implement the method of claim 1, characterized in that, The system includes: The data acquisition module is used to acquire the rated reactive power and rated excitation current of each generator set, and to determine the reactive power trend value. The benchmark screening module is used to obtain the relative deviation score of the excitation characteristics of each generator set based on the rated reactive power, rated excitation current and cumulative operating time of each key load point in each generator set; determine the operating time score based on the operating time of each generator set; and screen benchmark units by combining the relative deviation score of excitation characteristics and the operating time score. The judgment module is used to determine whether to start the adjustment process based on the deviation of the reactive power trend value of each generator set from that of the benchmark generator set, combined with the fluctuation characteristics of the reactive power trend value of each generator set. The control module is used to call the corresponding PI control parameters according to the load rate range when the control process is started, calculate the theoretical correction amount of reactive power deviation, limit the theoretical correction amount, and output the excitation current setpoint to the automatic voltage regulator of the corresponding generator set.