Variable frequency parameter regulation method for industrial control system

Abstract

The present application relates to the field of industrial process control system and programmed control technology, disclose a kind of variable frequency parameter regulation and control method for industrial control system, comprising: the process feedback of acquisition variable frequency control system, variable frequency output frequency, pulse width modulation duty cycle and frequency converter input power and calculate process deviation;When process deviation is in target interval, frequency and duty cycle micro disturbance are alternately applied, process response slope factor and power sensitivity factor are calculated;Cooperative cost function is established and regulation and control parameter reference point is locked;Decoupling is slow and double time domain loop, fast time domain loop is corrected duty cycle according to power sensitivity factor, and slow time domain loop is corrected variable frequency output frequency according to process deviation and process response slope factor, the present application is decoupled by double time domain, suppresses high-frequency power fluctuation and smooth tracking target parameter, reduces the contradiction between response overshoot and energy consumption swing, guarantees system operation safety.

Description

Technical Field

[0001] This invention belongs to the field of industrial process control system and programmed control technology, and particularly relates to a method for frequency conversion parameter control in industrial control systems. Background Technology

[0002] Variable frequency drive technology has been widely used in the operation control of power equipment such as industrial pumps, fans, compressors and conveyors. Common industrial variable frequency drive systems usually use pre-set proportional-integral-derivative control parameters or fixed voltage-frequency characteristic curves to collect process variables such as pressure and flow rate, calculate the deviation between the process variables and the target set value, and adjust the variable frequency output frequency and pulse width modulation duty cycle of the inverter accordingly to keep physical quantities such as pressure and flow rate stable in the production process.

[0003] In continuous industrial process control, the controlled object typically exhibits characteristics such as nonlinearity, large time delay, and load variation with operating conditions. When the system load changes suddenly or the operating conditions change, conventional control methods using fixed parameters often struggle to simultaneously achieve both response speed and stable operation. Increasing the proportional gain can accelerate the response, but it can also cause the inverter output frequency and pulse width modulation duty cycle to change too quickly, leading to process variable overshoot and repeated oscillations, resulting in significant fluctuations in pipeline pressure and even triggering protection circuit breaks. Adopting a more conservative parameter adjustment method can cause the inverter's response to lag significantly during the transition process, causing the motor to deviate from its efficient operating range and generate more ineffective power consumption. Furthermore, extensive testing of control parameters without constraints may exceed the stable operating range allowed by the industrial control system, causing process parameters to exceed limits.

[0004] Therefore, the technical problem that needs to be solved is how to perform programmed micro-amplitude probing through the main control unit of the industrial control system, under the premise of ensuring the operating range and steady-state safety of the industrial control system, identify the response characteristics of the industrial process online, adjust the pulse width modulation duty cycle and frequency conversion output frequency according to different time scales (fast and slow), and at the same time stop the probing in time and suppress DC bus voltage fluctuations when sudden interference occurs. Summary of the Invention

[0005] To address the problems in the background art, the technical solution of the present invention is as follows: A method for frequency converter parameter control in industrial control systems, comprising the following steps: Step S1: Collect the process feedback quantity, frequency converter output frequency, pulse width modulation duty cycle and frequency converter input power of the frequency converter control system. Calculate and generate the process deviation quantity that characterizes the system state deviation by subtracting the process feedback quantity from the preset process target parameters. Step S2: When the process deviation is within the preset target deviation range, frequency micro-disturbance and duty cycle micro-disturbance with amplitude lower than the rated set limit are alternately applied to the inverter drive stage within the preset first control cycle. Based on the process feedback and the rate of change of the inverter input power before and after the disturbance, the process response slope factor and power sensitivity factor are calculated respectively. Step S3: Based on the process deviation, inverter input power, process response slope factor and power sensitivity factor, establish a collaborative cost function that integrates control deviation and power loss, calculate the parameter correction along the negative gradient update direction of the collaborative cost function, and lock the optimal control parameter reference point. Step S4: Decouple the frequency converter control into a fast time-domain loop and a slow time-domain loop. The fast time-domain loop corrects the pulse width modulation duty cycle based on the power sensitivity factor in the second control cycle, which is shorter than the first control cycle, to suppress reactive power fluctuations. The slow time-domain loop corrects the frequency converter output frequency based on the process deviation and the process response slope factor in the third control cycle, which is longer than the second control cycle, to smoothly track the process target parameters. The first control cycle is not greater than the second control cycle.

[0006] Preferably, step S2 includes: step S21, during the first control cycle, alternately applying frequency micro-perturbations and duty cycle micro-perturbations with amplitudes lower than the preset rated amplitude limit to the inverter drive stage; step S22, calculating the process response slope factor based on the ratio of the change in process feedback quantity before and after applying the frequency micro-perturbation to the change in inverter output frequency; step S23, calculating the power sensitivity factor based on the ratio of the change in inverter input power before and after applying the duty cycle micro-perturbation to the change in pulse width modulation duty cycle.

[0007] Preferably, step S4 includes: step S41, in the second control cycle, the current pulse width modulation duty cycle is obtained by subtracting the product of the duty cycle step size coefficient and the power sensitivity factor from the pulse width modulation duty cycle of the previous cycle; step S42, in the third control cycle, the current frequency converter output frequency is obtained by subtracting the product of the frequency step size coefficient and the quotient of the process deviation divided by the process response slope factor from the frequency converter output frequency of the previous cycle.

[0008] Preferably, it also includes a trial blocking step S5: step S51, monitoring the change gradient of the process deviation; step S52, when the change gradient exceeds the preset deviation gradient threshold, clearing the frequency micro-perturbation and duty cycle micro-perturbation, locking the current frequency converter output frequency and pulse width modulation duty cycle, and reverting the frequency converter control parameters to the parameter point of the previous steady-state cycle.

[0009] Preferably, it also includes a feedforward compensation step S6: Step S61, extract the high-frequency ripple component of the inverter DC bus voltage; Step S62, when the high-frequency ripple component exceeds the preset ripple threshold, calculate the duty cycle feedforward compensation amount based on the high-frequency ripple component, and superimpose the duty cycle feedforward compensation amount onto the control output of the pulse width modulation duty cycle.

[0010] Preferably, the method further includes a dead time fine-tuning step S7: Step S71, obtain the current inverter output frequency and load current amplitude; Step S72, adjust the dead time of the inverter bridge switching transistor according to the inverter output frequency and load current amplitude, shorten the dead time under low frequency and low load conditions, and restore the standard dead time under high frequency and heavy load conditions.

[0011] Preferably, the trial blocking step S5 further includes: step S53, when the process deviation exceeds the preset target deviation range and the change gradient does not exceed the preset deviation gradient threshold, the micro-disturbance trial of the first control cycle is cut off, the frequency conversion control parameters are kept at the current average value, and the frequency conversion output frequency is adjusted according to the large deviation tracking mode.

[0012] Preferably, the fast time-domain loop and the slow time-domain loop respond in parallel during the full operating condition. The fast time-domain loop directly acts on the inverter bridge conduction state by finely adjusting the pulse width modulation duty cycle to suppress reactive power fluctuations, while the slow time-domain loop smoothly shifts the inverter output frequency along the process response slope to suppress DC bus voltage fluctuations.

[0013] Preferably, step S3 includes: step S31, calculating the parameter correction along the negative gradient update direction of the cooperative cost function, determining the optimal control parameter reference point of the frequency converter output frequency and the pulse width modulation duty cycle, and locking the initial parameters of the frequency converter control system at the optimal control parameter reference point when switching from the trial stage to the full operating condition stage.

[0014] Compared with existing technologies, the frequency conversion parameter control method of the present invention for industrial control systems has the following advantages: 1. In the frequency conversion parameter control of industrial control systems, the frequency conversion control parameters can be adjusted according to the actual response and operating power of the industrial process object, which improves the problem that fixed parameters are difficult to adapt to nonlinear operating conditions. The fast time domain loop makes a small adjustment to the pulse width modulation duty cycle, which can suppress the reactive power fluctuation caused by high-frequency load fluctuations in a timely manner. The slow time domain loop smoothly adjusts the frequency conversion output frequency, which can stably track the process target parameters and avoid process variable overshoot and oscillation caused by excessive frequency adjustment. Therefore, this method can reduce ineffective power consumption while ensuring process control accuracy, and achieve coordinated optimization of control response accuracy and operating energy consumption.

[0015] 2. It can promptly stop micro-disturbance testing when disturbances such as sudden load changes occur, and restore the frequency converter control parameters to the parameter points of the previous steady-state cycle, avoiding the superposition of testing actions and external disturbances. This can limit parameter testing to the stable operating range allowed by the industrial control system, reduce the risk of process parameter out-of-bounds, process variable overshoot, and protection circuit failure, and enable the frequency converter drive system to maintain safe and stable operation when the operating conditions change significantly, and improve the anti-interference capability of the control process.

[0016] 3. It can promptly perform duty cycle feedforward compensation when high-frequency ripple occurs in the DC bus voltage, suppress bus voltage fluctuations in advance when the process feedback has not changed significantly, and improve the response capability of the frequency converter control system to rapid electrical disturbances. At the same time, it can adjust the dead time of the inverter bridge switching transistors according to the frequency converter output frequency and load current amplitude, which can reduce output waveform distortion and switching heat loss under low frequency and low load conditions, and maintain the stable operation of the inverter bridge under high frequency and heavy load conditions, thus taking into account both the operating loss of the frequency converter and electrical safety. Attached Figure Description

[0017] Figure 1 This is a flowchart of the dual time-domain decoupling of the frequency conversion control parameters of the present invention; Figure 2 This is a diagram of the multi-dimensional frequency conversion collaborative control structure of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0019] A method for frequency converter parameter control in industrial control systems includes the following steps: Step S1: Collect the process feedback quantity, frequency converter output frequency, pulse width modulation duty cycle and frequency converter input power of the frequency converter control system. Calculate and generate the process deviation quantity that characterizes the system state deviation by subtracting the process feedback quantity from the preset process target parameters. Step S2: When the process deviation is within the preset target deviation range, frequency micro-disturbance and duty cycle micro-disturbance with amplitude lower than the rated set limit are alternately applied to the inverter drive stage within the preset first control cycle. Based on the process feedback and the rate of change of the inverter input power before and after the disturbance, the process response slope factor and power sensitivity factor are calculated respectively. Step S3: Based on the process deviation, inverter input power, process response slope factor and power sensitivity factor, establish a collaborative cost function that integrates control deviation and power loss, calculate the parameter correction along the negative gradient update direction of the collaborative cost function, and lock the optimal control parameter reference point. Step S4: Decouple the frequency converter control into a fast time-domain loop and a slow time-domain loop. The fast time-domain loop corrects the pulse width modulation duty cycle based on the power sensitivity factor in the second control cycle, which is shorter than the first control cycle, to suppress reactive power fluctuations. The slow time-domain loop corrects the frequency converter output frequency based on the process deviation and the process response slope factor in the third control cycle, which is longer than the second control cycle, to smoothly track the process target parameters. The first control cycle is not greater than the second control cycle.

[0020] Preferably, step S2 includes: step S21, during the first control cycle, alternately applying frequency micro-perturbations and duty cycle micro-perturbations with amplitudes lower than the preset rated amplitude limit to the inverter drive stage; step S22, calculating the process response slope factor based on the ratio of the change in process feedback quantity before and after applying the frequency micro-perturbation to the change in inverter output frequency; step S23, calculating the power sensitivity factor based on the ratio of the change in inverter input power before and after applying the duty cycle micro-perturbation to the change in pulse width modulation duty cycle.

[0021] Preferably, step S4 includes: step S41, in the second control cycle, the current pulse width modulation duty cycle is obtained by subtracting the product of the duty cycle step size coefficient and the power sensitivity factor from the pulse width modulation duty cycle of the previous cycle; step S42, in the third control cycle, the current frequency converter output frequency is obtained by subtracting the product of the frequency step size coefficient and the quotient of the process deviation divided by the process response slope factor from the frequency converter output frequency of the previous cycle.

[0022] Preferably, it also includes a trial blocking step S5: step S51, monitoring the change gradient of the process deviation; step S52, when the change gradient exceeds the preset deviation gradient threshold, clearing the frequency micro-perturbation and duty cycle micro-perturbation, locking the current frequency converter output frequency and pulse width modulation duty cycle, and reverting the frequency converter control parameters to the parameter point of the previous steady-state cycle.

[0023] Preferably, it also includes a feedforward compensation step S6: Step S61, extract the high-frequency ripple component of the inverter DC bus voltage; Step S62, when the high-frequency ripple component exceeds the preset ripple threshold, calculate the duty cycle feedforward compensation amount based on the high-frequency ripple component, and superimpose the duty cycle feedforward compensation amount onto the control output of the pulse width modulation duty cycle.

[0024] Preferably, the method further includes a dead time fine-tuning step S7: Step S71, obtain the current inverter output frequency and load current amplitude; Step S72, adjust the dead time of the inverter bridge switching transistor according to the inverter output frequency and load current amplitude, shorten the dead time under low frequency and low load conditions, and restore the standard dead time under high frequency and heavy load conditions.

[0025] Preferably, the trial blocking step S5 further includes: step S53, when the process deviation exceeds the preset target deviation range and the change gradient does not exceed the preset deviation gradient threshold, the micro-disturbance trial of the first control cycle is cut off, the frequency conversion control parameters are kept at the current average value, and the frequency conversion output frequency is adjusted according to the large deviation tracking mode.

[0026] Preferably, the fast time-domain loop and the slow time-domain loop respond in parallel during the full operating condition. The fast time-domain loop directly acts on the inverter bridge conduction state by finely adjusting the pulse width modulation duty cycle to suppress reactive power fluctuations, while the slow time-domain loop smoothly shifts the inverter output frequency along the process response slope to suppress DC bus voltage fluctuations.

[0027] Preferably, step S3 includes: step S31, calculating the parameter correction along the negative gradient update direction of the cooperative cost function, determining the optimal control parameter reference point of the frequency converter output frequency and the pulse width modulation duty cycle, and locking the initial parameters of the frequency converter control system at the optimal control parameter reference point when switching from the trial stage to the full operating condition stage.

[0028] Example 1: This example is applied to an industrial water pump control system. Its variable frequency drive has a rated power of 37kW, a rated voltage of 380V, and a rated frequency of 50Hz. The pulse width modulation carrier frequency is 4kHz. A water pressure sensor collects the water pressure in the pipeline and uses it as a process feedback quantity. Process target parameters The pressure is set to 0.40 MPa, the sampling period of the frequency converter is set to 10 ms, and the process feedback quantity is measured when the downstream pipeline valves open and close rapidly. With process target parameters The process deviation is generated in real time. Its calculation formula is When the pipeline pressure dynamically drifts, increasing the proportional gain can easily cause water pressure overshoot oscillation and lead to pipeline pressure overpressure protection circuit breaking. When conservative adjustment is used, the frequency converter response is lagging, the motor deviates from the high-efficiency operating range, and reactive power fluctuations are generated at the input of the frequency converter.

[0029] At any moment That is, at 100ms, the water pressure in the pipeline is 0.405MPa, and the frequency converter output frequency is... 49.80Hz, pulse width modulation duty cycle The inverter input power is 0.650. The value is 15200W, from which the process deviation is calculated. The process deviation is 0.005 MPa. The absolute value is lower than the half-width of the preset target deviation range. That is, at 0.01 MPa, the system enters a steady-state micro-amplitude trial state, and the frequency converter is at time [time missing]. That is, a frequency perturbation with an amplitude of 0.20Hz is applied to the inverter drive stage at 110ms. The inverter output frequency was temporarily adjusted to 50.00Hz; at this time, the collected pipeline water pressure was 0.408MPa, the water pressure change was 0.003MPa, and the inverter input power became 15350W. Based on the feedback before and after the test... The change in frequency and the frequency converter output frequency The ratio of the changes is used to calculate the process response slope factor. It is 0.015 MPa / Hz, and then, at time... That is, a duty cycle perturbation with an amplitude of 0.01 is applied at 120ms. Based on the input power of the frequency converter before and after the test Variation and Pulse Width Modulation Duty Cycle The power sensitivity factor is calculated from the ratio of the changes. 15kW / unit.

[0030] When constructing the collaborative cost function, the process deviation is used. The squared term characterizes the control deviation, expressed as the inverter input power. Characterizes the power loss in the variable frequency drive process, and is expressed by the process response slope factor. and power sensitivity factor Determine the correction direction for the inverter output frequency and the pulse width modulation duty cycle, respectively, and the error weighting coefficient. Take 100MPa -2 Power loss weighting coefficient Take 0.001W -1 Cooperative cost function The expression is The process deviation of 0.005 MPa and the time Substituting the corresponding inverter input power of 15350W, we obtain the value of the collaborative cost function. The value is 15.3525. When calculating along the negative gradient update direction of the collaborative cost function, the duty cycle direction is determined based on the power sensitivity factor. It is determined that the correction amount for the pulse width modulation duty cycle is the product of the duty cycle step size factor and the power sensitivity factor; the frequency direction is determined based on the process deviation. With process response slope factor The ratio is determined, and the correction amount of the variable frequency output frequency is the product of the frequency step size coefficient and this ratio, based on the process response slope factor. With power sensitivity factor The value of the cooperative cost function is recursively updated along the negative gradient direction. After each correction of the frequency converter output frequency and pulse width modulation duty cycle, the value of the cooperative cost function is recalculated. When continuing to update along the negative gradient direction no longer reduces the value of the cooperative cost function, the combination of the frequency converter output frequency and pulse width modulation duty cycle of the previous control cycle is determined as the optimal control parameter reference point. When switching from the trial stage to the full-condition operation stage, the initial parameters of the frequency converter control system are locked at this optimal control parameter reference point.

[0031] Based on this, the frequency converter control is decoupled into a fast time-domain loop and a slow time-domain loop. In the second control cycle, the fast time-domain loop subtracts the duty cycle step size coefficient from the pulse width modulation duty cycle of the previous cycle. With power sensitivity factor The product of these factors yields the current pulse width modulation duty cycle and the duty cycle step size coefficient. Take 0.0005kW -1 The time was calculated That is, the pulse width modulation duty cycle at 130ms The power sensitivity factor is 0.6425. Reflecting the impact of duty cycle changes on the inverter input power, the fast time-domain loop adjusts the pulse width modulation duty cycle accordingly, changing the inverter bridge conduction state and suppressing reactive power fluctuations at the inverter input. The slow time-domain loop, within the third control cycle, subtracts the frequency step size coefficient from the inverter output frequency of the previous cycle. process deviation Divide by the process response slope factor The product of the obtained ratios yields the current inverter output frequency and the frequency step size coefficient. Using 0.05 Hz / MPa, the time was calculated. The frequency converter output frequency at the location The frequency is 49.78 Hz, which is used to smoothly track the process target parameters corresponding to the pipeline water pressure.

[0032] At any moment At 200ms, the pipeline valve rapidly closed, and the pipeline water pressure rose to 0.435MPa. Within a 10ms sampling interval, the water pressure change was 0.030MPa, and the gradient of the process deviation was 0.003MPa / ms. This gradient did not exceed the preset deviation gradient threshold. That is, 0.005 MPa / ms, but the process deviation is... The absolute value reached 0.035 MPa, exceeding the preset target deviation range. The frequency converter switched to a micro-amplitude trial state and cleared the frequency micro-disturbance. and duty cycle perturbation Before entering the large deviation tracking mode, the average values ​​of the inverter output frequency and pulse width modulation duty cycle samples in the previous steady-state cycle are calculated, and this set of average values ​​is used as the current average value state. After switching out of the micro-amplitude probing mode, the pulse width modulation duty cycle is maintained at the current average value state, and the inverter output frequency is corrected according to the large deviation tracking mode. The frequency correction factor is a negative frequency step size coefficient. process deviation Divide by the process response slope factor The product of the obtained ratios yields a frequency adjustment of -0.1167Hz. The inverter output frequency will then be adjusted accordingly. The frequency was lowered to 49.66Hz to reduce pipeline water pressure overshoot, so that the industrial water pump control system could remain within the physical operating boundary during sudden load changes and suppress reactive power fluctuations.

[0033] The switching rule for the large deviation tracking mode is as follows: when the absolute value of the process deviation is greater than the half-width of the preset target deviation range (0.01 MPa) and the gradient of the process deviation does not exceed the preset deviation gradient threshold (0.005 MPa / ms), the system switches to the large deviation tracking mode. In this mode, the fast time-domain loop pauses the micro-step correction based on the power sensitivity factor, and the pulse width modulation duty cycle remains at the current average value. The slow time-domain loop shortens the third control cycle from 50 ms to 20 ms, and simultaneously amplifies the frequency step coefficient from 0.05 Hz / MPa by 2.5 times to 0.125 Hz / MPa to improve the update speed of the inverter output frequency. After the process deviation re-enters the preset target deviation range, the system resumes the control state of parallel operation of the fast time-domain loop and the slow time-domain loop.

[0034] Example 2: This example establishes an industrial water pump frequency converter drive test platform to test its operating status under dynamic load disturbance and high-frequency electromagnetic interference conditions. The test platform's frequency converter drive has a rated power of 37kW, a rated voltage of 380V, a rated frequency of 50Hz, and a pulse width modulation carrier frequency of 4kHz. The water pressure sensor has a measurement range of 0 to 1.00MPa and a resolution of 0.001MPa. The power measurement accuracy of the electric force measurement unit is 0.1%, and the sampling period is set to 10ms. Gaussian white noise with a signal-to-noise ratio of 20dB and 50Hz power frequency harmonic interference are superimposed on the transmission line of the process feedback quantity to simulate the electromagnetic noise in the industrial field.

[0035] The sampling period is determined based on the dynamic response bandwidth of the pipeline water pressure and the floating-point operation load of the frequency converter. The main frequency energy of the pipeline water pressure fluctuation is concentrated below 5Hz. According to the Nyquist sampling requirements and with an oversampling margin of 5 to 10 times, the sampling frequency is set to 100Hz, and the corresponding sampling period is 10ms. The amplitude of the frequency micro-disturbance is set to 0.20Hz, accounting for 0.4% of the rated frequency of 50Hz; the amplitude of the duty cycle micro-disturbance is set to 0.01, accounting for 1% of the rated duty cycle. The above amplitudes can make the calculated signals of the process response slope factor and power sensitivity factor higher than the background noise, while limiting the micro-disturbance energy and avoiding water hammer effect in the pipeline.

[0036] The duty cycle feedforward compensation is 0.00067V. -1 The feedforward compensation coefficient is calculated based on the charge-discharge energy relationship under the conditions of a DC bus capacitance of 1000μF and a rated DC voltage of 540V. The duty cycle step size coefficient is set to 0.0005kW. -1 The frequency step size coefficient was set to 0.05 Hz / MPa. The above two step size coefficients were determined through an offline simulation platform. The simulation object was a 37kW water pump system. 50 sets of step response tests were completed under positive and negative 20% rated load disturbances. The parameters were selected based on shortening the response time and reducing reactive power fluctuations. The preset deviation gradient threshold was set to 0.005MPa / ms, which corresponds to a 50% safety margin of the test limit value of the maximum water hammer pressure rise rate of the water pump.

[0037] During the test, a power quality analyzer was used to continuously collect the discrete sequences of instantaneous voltage and instantaneous current at the input of the frequency converter at a sampling rate of 10kHz. 100 sampling points were selected in each complete power frequency cycle to calculate the discrete value of reactive power. The difference between the maximum peak value and the minimum valley value of reactive power in 10 consecutive power frequency cycles was taken as the reactive power fluctuation. The recovery time was recorded as the time required for the process deviation to re-enter the preset target deviation range after the disturbance was applied.

[0038] The experiment included four control groups: sample group, control group 1, control group 2, control group 3, and control group 4. The target process parameter was set to 0.40 MPa. Control group 1 used fixed proportional-integral-derivative control; control group 2 did not use decoupling control between the fast and slow time domain loops, and instead used a single-cycle synchronous update of the inverter output frequency and pulse width modulation duty cycle; control group 3 amplified the frequency perturbation to 2.50 Hz, i.e., 5% of the rated frequency; and control group 4 removed the trial blocking step executed based on the gradient of process deviation changes. During the test, the pipeline valve was suddenly opened or closed... To handle sudden load changes, the system collects process feedback, inverter output frequency, pulse width modulation duty cycle, and inverter input power in real time. In this embodiment, when the process deviation enters the preset target deviation range, frequency micro-perturbation and duty cycle micro-perturbation are applied alternately in the form of square waves. The process feedback superimposed with 20dB Gaussian white noise is processed using time-domain differential smoothing filter, and the process response slope factor and power sensitivity factor are calculated. The fast time-domain loop corrects the pulse width modulation duty cycle according to the second control cycle of 10ms, and the slow time-domain loop corrects the inverter output frequency according to the third control cycle of 50ms.

[0039] In steady-state condition number 1, the process feedback collected by the sample group in this embodiment is 0.403 MPa. The difference between this value and the process target parameter of 0.40 MPa yields a process deviation of 0.003 MPa. Based on the process feedback before and after the micro-disturbance, the inverter output frequency, the inverter input power, and the pulse width modulation duty cycle, a process response slope factor of 0.015 MPa / Hz and a power sensitivity factor of 15.1 kW / unit are calculated. After parameter corrections are performed in the fast time domain loop and the slow time domain loop respectively, the inverter output frequency is 49.82 Hz, the pulse width modulation duty cycle is 0.648, the recovery time is 120 ms, and the reactive power fluctuation is 18 W.

[0040] In the micro-disturbance condition No. 2, after the frequency micro-disturbance and duty cycle micro-disturbance are applied, the process feedback is 0.408 MPa, the corresponding process deviation is 0.008 MPa, the process response slope factor calculated from the data before and after the disturbance is 0.015 MPa / Hz, and the power sensitivity factor is 14.9 kW / unit; the frequency of the inverter output after parameter correction is 50.00 Hz, the pulse width modulation duty cycle is 0.650, the recovery time is 130 ms, and the reactive power fluctuation is 22 W.

[0041] In valve closure condition #3, the process feedback value collected after the pipeline water pressure rises is 0.432 MPa, and the corresponding process deviation is 0.032 MPa. At this time, the process response slope factor is 0.015 MPa / Hz, and the power sensitivity factor is 15.0 kW / unit. After the process deviation exceeds the preset target deviation range, the system switches to micro-disturbance probing, maintains the current average value of the pulse width modulation duty cycle, and adjusts the frequency converter output frequency according to the large deviation tracking mode. The adjusted frequency converter output frequency is 49.68 Hz, the pulse width modulation duty cycle is 0.641, and the process deviation re-enters the preset target deviation range after 180 ms. The reactive power fluctuation is 25 W.

[0042] In the bus ripple impact condition No. 4, the collected process feedback is 0.412 MPa, the corresponding process deviation is 0.012 MPa, the process response slope factor is 0.014 MPa / Hz, and the power sensitivity factor is 15.2 kW / unit. The frequency converter extracts the high-frequency ripple component of the DC bus voltage, calculates the duty cycle feedforward compensation based on the high-frequency ripple component, and superimposes it on the control output of the pulse width modulation duty cycle. After compensation, the frequency converter output frequency is 49.75 Hz, the pulse width modulation duty cycle is 0.643, the recovery time is 140 ms, and the reactive power fluctuation is 20 W.

[0043] In control group 1, the fixed proportional-integral-derivative (PID) control does not apply frequency or duty cycle perturbations, therefore the process response slope factor and power sensitivity factor are not calculated and are both denoted as --. Number 5 corresponds to the steady-state condition, with a process feedback of 0.406 MPa, a process deviation of 0.006 MPa, a frequency converter output of 49.85 Hz, a pulse width modulation duty cycle of 0.652, a recovery time of 480 ms, and a reactive power fluctuation of 165 W. Number 6 corresponds to the valve sudden closure condition, where the fixed control parameters are not adjusted online according to load changes, the process feedback increases to 0.448 MPa, and the process deviation reaches 0.048 MPa; at this time, the frequency converter output is 49.20 Hz, the pulse width modulation duty cycle is 0.620, the recovery time is extended to 850 ms, and the reactive power fluctuation increases to 320 W.

[0044] Number 7 represents the single-time-domain update condition of control group 2. The process feedback is 0.415 MPa, the corresponding process deviation is 0.015 MPa, the process response slope factor is 0.015 MPa / Hz, and the power sensitivity factor is 15.0 kW / unit. Since the inverter output frequency and the pulse width modulation duty cycle are synchronously corrected in the same cycle, the two control quantities affect each other. The corrected inverter output frequency is 49.70 Hz, the pulse width modulation duty cycle is 0.635, the recovery time is 340 ms, and the reactive power fluctuation is 142 W.

[0045] Number 8 represents the out-of-range micro-perturbation condition of control group 3. After amplifying the frequency micro-perturbation to 2.50Hz, the process feedback is 0.428MPa, the process deviation is 0.028MPa, the calculated process response slope factor decreases to 0.008MPa / Hz, and the power sensitivity factor increases to 28.5kW / unit. The frequency micro-perturbation has a significant impact on the steady-state hydraulic state of the pump flow channel. The frequency converter output frequency increases to 52.30Hz, the pulse width modulation duty cycle is 0.680, the recovery time is 620ms, and the reactive power fluctuation is 280W. Compared with the 0.20Hz frequency micro-perturbation, the 2.50Hz frequency micro-perturbation amplifies the process deviation and reactive power fluctuation.

[0046] Number 9 represents the gradient-free blocking condition of control group 4. When the pipeline valve is suddenly closed, the gradient of the process deviation reaches 0.003 MPa / ms. Since the trial blocking step is removed, the frequency and duty cycle perturbations continue to act in conjunction with the external load impact, causing the process feedback to rise to 0.452 MPa, corresponding to a process deviation of 0.052 MPa. At this point, the process response slope factor is 0.015 MPa / Hz, the power sensitivity factor is 15.0 kW / unit, and the inverter output frequency drops to 49 Hz. With a frequency of 0.10Hz, a pulse width modulation duty cycle of 0.615, a recovery time of 920ms, and a reactive power fluctuation of 390W, when the system retains the trial blocking and large deviation tracking control, it can stop the micro-disturbance trial after the process deviation exceeds the preset target deviation range, and the slow time domain loop continues to adjust the frequency converter output frequency. When the bus ripple impact occurs, the duty cycle feedforward compensation is directly superimposed on the control output of the pulse width modulation duty cycle, thereby limiting the impact of load mutation and bus ripple on the frequency converter control process.

[0047] Example 3: This example uses an industrial water pump control system with a rated power of 37kW, a rated voltage of 380V, a rated frequency of 50Hz, and a pulse width modulation carrier frequency of 4kHz. A water pressure sensor collects the pipeline water pressure as the process feedback quantity. Process target parameters When the pressure is set to 0.40 MPa, the sampling period is set to 10 ms, and 20 dB Gaussian white noise and 50 Hz power frequency harmonic interference are superimposed on the transmission line of the process feedback, the frequency converter uses time-domain differential smoothing filtering to process the collected pipeline water pressure and then uses the processed process feedback quantity... With process target parameters Calculate process deviation .

[0048] Based on the dynamic response bandwidth of the pipeline water pressure and the operational load of the frequency converter, the half-width of the preset target deviation range is determined. Set to 0.01 MPa, and set the preset deviation gradient threshold. Set to 0.005 MPa / ms, when the process deviation... When the absolute value is not greater than 0.01 MPa, the system enters a steady-state micro-disturbance probing state, and during the first control cycle, frequency micro-disturbances with an amplitude of 0.20 Hz are alternately applied to the inverter drive stage. And a duty cycle perturbation with an amplitude of 0.01 The amplitude of the frequency micro-perturbation is 0.4% of the rated frequency of 50Hz, and the amplitude of the duty cycle micro-perturbation is 1% of the rated duty cycle. This allows the changes in process feedback and inverter input power before and after the perturbation to be distinguished from the background noise, while limiting the impact of the micro-perturbation on the pipeline water pressure. The inverter controller calculates the process response slope factor based on the changes in process feedback and inverter output frequency before and after the frequency micro-perturbation. The power sensitivity factor is calculated based on the changes in the inverter input power and pulse width modulation duty cycle before and after the duty cycle micro-perturbation. .

[0049] The first control cycle is set to 10ms, which is used to apply micro-disturbances and collect the corresponding process feedback and inverter input power. The second control cycle is also set to 10ms, which is used for the fast time-domain loop to update the pulse width modulation duty cycle. The third control cycle is set to 50ms, which is used for the slow time-domain loop to update the inverter output frequency. The micro-disturbance sampling and the fast time-domain loop operate under the same 10ms control clock, and the slow time-domain loop performs a frequency correction every 50ms.

[0050] When rapid opening and closing of pipeline valves or random load shocks cause sudden changes in pipeline water pressure, the frequency converter calculates the process deviation in real time. The gradient of change, when the gradient of change exceeds the preset deviation gradient threshold. At that time, the frequency perturbation was cleared. and duty cycle perturbation Lock the current inverter output frequency With pulse width modulation duty cycle The frequency converter control parameters are then rolled back to the parameter points of the previous steady-state cycle to prevent the superposition of micro-disturbance testing and the current load impact. When the process deviation... When the absolute value exceeds 0.01 MPa and its change gradient does not exceed 0.005 MPa / ms, the frequency converter switches off the micro-disturbance probe mode, maintains the frequency converter control parameters at the current average value, and adjusts the frequency converter output frequency according to the large deviation tracking mode. In this mode, according to the frequency step size coefficient process deviation Divide by the process response slope factor The product of the obtained ratios corrects the frequency converter output frequency, causing the process deviation to re-enter the preset target deviation range.

[0051] During full-condition operation, the fast time-domain loop and the slow time-domain loop respond in parallel. In the second control cycle, the fast time-domain loop subtracts the duty cycle step size coefficient from the pulse width modulation duty cycle of the previous cycle. With power sensitivity factor The product of these two factors gives the current pulse width modulation duty cycle. Furthermore, reactive power fluctuations are suppressed by changing the conduction state of the inverter bridge. In the third control cycle, the slow-time domain loop subtracts the frequency step size coefficient from the inverter output frequency of the previous cycle. process deviation Divide by the process response slope factor The product of the obtained ratios gives the current inverter output frequency. This enables the pipeline water pressure to smoothly track the target parameters during the process. .

[0052] The slow-time domain loop smoothly adjusts the inverter output frequency, impacting the hydraulic mechanical inertia of the pump and pipe network. This limits sudden changes in pump shaft power and reduces low-frequency energy fed back to the DC bus capacitor during motor braking or sudden load increases. For DC bus voltage changes in the microsecond to millisecond range, the fast-time domain loop responds by adjusting the inverter bridge's conduction state. The inverter controller continuously samples the DC bus voltage, subtracts the slowly varying average component from the sampled value, and uses the remaining variation component as the high-frequency ripple component. When the high-frequency ripple component Exceeding the preset ripple threshold At that time, the duty cycle feedforward compensation amount is calculated based on the high-frequency ripple component. And superimposed on the pulse width modulation duty cycle In the control output.

[0053] The frequency converter also adjusts according to the current frequency output frequency. The dead time of the upper and lower arm switching transistors of the inverter bridge is adjusted according to the load current amplitude. The system is set as follows: 30% of the rated frequency of 50Hz, i.e., 15Hz, is set as the low-frequency threshold, and 40% of the rated motor current is set as the low-load threshold. When the inverter output frequency is lower than 15Hz and the load current amplitude is lower than 40% of the rated current, the system is in a low-frequency, low-load condition. The dead time is shortened from the standard dead time of 2.5μs to 1.5μs to reduce waveform zero-crossing distortion and switching losses under low current conditions. When the inverter output frequency is not lower than 15Hz and the load current amplitude is not lower than 40% of the rated current, the system is in a high-frequency, heavy-load condition. The dead time is restored to 2.5μs to maintain the shoot-through protection time between the upper and lower arms of the inverter bridge.

[0054] Example 4: This example combines Figures 1 to 2 This section explains the frequency converter parameter control methods used in industrial control systems, such as... Figure 1 As shown, in step S1, the process feedback quantity, inverter output frequency, pulse width modulation duty cycle, and inverter input power of the frequency converter control system are collected. The process deviation quantity, which characterizes the system state deviation, is calculated by subtracting the process feedback quantity from the preset process target parameter. In step S2, when the process deviation quantity is within the preset target deviation range, frequency micro-perturbations and duty cycle micro-perturbations with amplitudes lower than the rated set limit are alternately applied to the inverter drive stage within the preset first control cycle. Based on the change rate of the process feedback quantity and the inverter input power before and after the perturbation, the process response slope factor and power sensitivity factor are calculated respectively. In step S3, based on the process deviation quantity and the inverter input power, the process response slope factor and power sensitivity factor are calculated respectively. A collaborative cost function for integrating control deviation and power loss is established using input power, process response slope factor, and power sensitivity factor. The parameter correction amount is calculated along the negative gradient update direction of the collaborative cost function to lock the optimal control parameter reference point. In step S4, the frequency converter control is decoupled into a fast time domain loop and a slow time domain loop. The fast time domain loop corrects the pulse width modulation duty cycle according to the power sensitivity factor in the second control cycle, which is shorter than the first control cycle, to suppress reactive power fluctuations. The slow time domain loop corrects the frequency converter output frequency according to the process deviation and process response slope factor in the third control cycle, which is longer than the second control cycle, to smoothly track the process target parameter. The first control cycle is not greater than the second control cycle.

[0055] like Figure 2 As shown, the frequency converter parameter control method for industrial control systems includes characterizing system state deviation, alternating application of micro-perturbations, fast and slow dual-time domain loops, and trial blocking and feedforward compensation. Characterizing system state deviation includes acquiring process feedback quantities and calculating process deviation quantities. Alternating application of micro-perturbations includes calculating power sensitivity factors and calculating process response slope factors. The fast and slow dual-time domain loops include the fast time domain loop suppressing reactive power fluctuations and the slow time domain loop smoothing and tracking process target parameters. Trial blocking and feedforward compensation includes superimposing duty cycle feedforward compensation quantities and clearing frequency micro-perturbations and duty cycle micro-perturbations.

[0056] Example 5: Before the industrial water pump control system is put into operation, a 24V DC power supply is first connected to the pipeline water pressure sensor and the frequency converter control module, and preheated for 15 minutes. When there is no medium flowing in the pipeline, the zero-point bias voltage of the water pressure sensor is collected, the zero-point output is calibrated to 0.000V, and this state is used as the measurement zero point of the process feedback quantity. Then, standard pressure medium is filled into the pipeline to make the pipeline pressure reach the process target parameter of 0.40MPa, and the full-scale output gain of the water pressure sensor is calibrated accordingly.

[0057] Before the inverter bridge triggers the pulse width modulation drive signal, the sampling unit extracts the high-frequency ripple component of the DC bus voltage and stores it in the static memory as the background reference when comparing the high-frequency ripple component with the 15V preset ripple threshold. After the pipeline impedance changes, the frequency converter continuously collects the pipeline water pressure and frequency converter input power for 50 cycles according to a sampling period of 10ms, calculates their average values, and uses this set of average values ​​as the initial state for steady-state determination, so that the process deviation obtained in subsequent calculations corresponds to the degree of deviation of the pipeline pressure from the process target parameter.

[0058] After 1000 hours of continuous operation, mechanical wear causes drift in the pump flow characteristics. The frequency converter initiates periodic online parameter updates. When the absolute value of the process deviation is no greater than 0.01 MPa for 100 consecutive sampling cycles, a frequency micro-perturbation with an amplitude of 0.20 Hz and a duty cycle micro-perturbation with an amplitude of 0.01 are alternately applied to the frequency converter drive stage during the first control cycle. The process response slope factor is updated based on the changes in the process feedback and the frequency converter output frequency before and after the frequency micro-perturbation; the power sensitivity factor is updated based on the changes in the frequency converter input power and the pulse width modulation duty cycle before and after the duty cycle micro-perturbation.

[0059] The frequency converter recalculates the parameter correction based on the updated process response slope factor and power sensitivity factor, and updates the optimal control parameter reference point. The fast time-domain loop is set to 0.0005kW. -1 The duty cycle step size coefficient corrects the pulse width modulation duty cycle, and the slow time domain loop corrects the frequency converter output frequency according to the frequency step size coefficient of 0.05Hz / MPa, so that the pipeline water pressure continues to track the process target parameters and suppresses the frequency converter input power fluctuation caused by the drift of the flow channel characteristics.

[0060] The embodiments of this application have been described above with reference to the accompanying drawings. Unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other. This application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit of this application and the scope of protection of this invention, and all of these forms are within the protection scope of this application.

Claims

1. A method for frequency conversion parameter control in industrial control systems, characterized in that, Includes the following steps: Step S1: Collect the process feedback quantity, frequency converter output frequency, pulse width modulation duty cycle and frequency converter input power of the frequency converter control system. Calculate and generate the process deviation quantity that characterizes the system state deviation by subtracting the process feedback quantity from the preset process target parameters. Step S2: When the process deviation is within the preset target deviation range, frequency micro-disturbance and duty cycle micro-disturbance with amplitudes lower than the rated set limit are alternately applied to the inverter drive stage within the preset first control cycle. Based on the process feedback and the rate of change of the inverter input power before and after the disturbance, the process response slope factor and power sensitivity factor are calculated respectively. Step S3: Based on the process deviation, inverter input power, process response slope factor and power sensitivity factor, establish a collaborative cost function that integrates control deviation and power loss, calculate the parameter correction along the negative gradient update direction of the collaborative cost function, and lock the optimal control parameter reference point. Step S4: Decouple the frequency converter control into a fast time-domain loop and a slow time-domain loop. The fast time-domain loop corrects the pulse width modulation duty cycle based on the power sensitivity factor in the second control cycle, which is shorter than the first control cycle, to suppress reactive power fluctuations. The slow time-domain loop corrects the frequency converter output frequency based on the process deviation and the process response slope factor in the third control cycle, which is longer than the second control cycle, to smoothly track the process target parameters. The first control cycle is not greater than the second control cycle.

2. The method for frequency conversion parameter control in an industrial control system according to claim 1, characterized in that, Step S2 includes: Step S21, during the first control cycle, alternately applying frequency micro-perturbations and duty cycle micro-perturbations with amplitudes lower than the preset rated amplitude limit to the inverter drive stage; Step S22, calculating the process response slope factor based on the ratio of the change in process feedback quantity before and after applying the frequency micro-perturbation to the change in inverter output frequency; Step S23, calculating the power sensitivity factor based on the ratio of the change in inverter input power before and after applying the duty cycle micro-perturbation to the change in pulse width modulation duty cycle.

3. The method for frequency conversion parameter control in an industrial control system according to claim 1, characterized in that, Step S4 includes: Step S41, in the second control cycle, the current pulse width modulation duty cycle is obtained by subtracting the product of the duty cycle step size coefficient and the power sensitivity factor from the pulse width modulation duty cycle of the previous cycle; Step S42, in the third control cycle, the current frequency converter output frequency is obtained by subtracting the product of the frequency step size coefficient and the quotient of the process deviation divided by the process response slope factor from the frequency converter output frequency of the previous cycle.

4. The method for frequency conversion parameter control in an industrial control system according to claim 1, characterized in that, It also includes a trial blocking step S5: Step S51, monitor the gradient of the process deviation; Step S52, when the gradient exceeds the preset deviation gradient threshold, clear the frequency micro-perturbation and duty cycle micro-perturbation, lock the current frequency converter output frequency and pulse width modulation duty cycle, and roll back the frequency converter control parameters to the parameter point of the previous steady state cycle.

5. The method for frequency conversion parameter control in an industrial control system according to claim 1, characterized in that, It also includes a feedforward compensation step S6: Step S61, extract the high-frequency ripple component of the inverter DC bus voltage; Step S62, when the high-frequency ripple component exceeds the preset ripple threshold, calculate the duty cycle feedforward compensation amount based on the high-frequency ripple component, and superimpose the duty cycle feedforward compensation amount onto the control output of the pulse width modulation duty cycle.

6. The method for frequency conversion parameter control in an industrial control system according to claim 1, characterized in that, It also includes a dead time fine-tuning step S7: Step S71, obtain the current inverter output frequency and load current amplitude; Step S72: Adjust the dead time of the inverter bridge switching transistors according to the inverter output frequency and load current amplitude. Shorten the dead time under low frequency and low load conditions, and restore the standard dead time under high frequency and heavy load conditions.

7. The method for frequency conversion parameter control in an industrial control system according to claim 4, characterized in that, The trial blocking step S5 also includes: step S53, when the process deviation exceeds the preset target deviation range and the change gradient does not exceed the preset deviation gradient threshold, the micro-disturbance trial of the first control cycle is cut off, the frequency conversion control parameters are kept at the current average value, and the frequency conversion output frequency is adjusted according to the large deviation tracking mode.

8. The method for frequency conversion parameter control in an industrial control system according to claim 1, characterized in that, The fast time-domain loop and the slow time-domain loop respond in parallel during the full operating condition. The fast time-domain loop directly acts on the inverter bridge conduction state by finely adjusting the pulse width modulation duty cycle to suppress reactive power fluctuations, while the slow time-domain loop smoothly shifts the inverter output frequency along the process response slope to suppress DC bus voltage fluctuations.

9. A method for frequency conversion parameter control in an industrial control system according to claim 1, characterized in that, Step S3 includes: Step S31, calculating the parameter correction along the negative gradient update direction of the cooperative cost function, determining the optimal control parameter reference point for the frequency converter output frequency and the pulse width modulation duty cycle, and locking the initial parameters of the frequency converter control system at the optimal control parameter reference point when switching from the trial phase to the full operating phase.

Images