Integrated variable frequency drive digital pump adaptive control method

Abstract

The application discloses a kind of integrated variable frequency drive's digital pump adaptive control method, belong to intelligent control technical field, comprising: receiving operating gear instruction and target control parameter, while collecting inlet and outlet temperature of conveying medium, real-time operating frequency;Based on the heat exchange optimal temperature difference deduction of conveying medium, the temperature deviation of conveying medium, flow matching deviation are calculated respectively;Through the pre-prediction of pump's front speed, the feedforward reference speed of pump is generated;Double-path PID speed regulation loop is constructed, and feedback operation is carried out, to generate modified reference speed;Through the weighted summation of feedforward reference speed and modified reference speed, the control speed of pump is obtained, and the control speed is limited amplitude processing, to obtain target speed and issue;Based on the continuous operation of pump motor, the inlet and outlet temperature difference of conveying medium is updated cyclically, and the steady-state cumulative deviation is calculated based on time sequence cache data to carry out fine tuning compensation determination, to generate compensation speed and superimposed to target speed.

Description

Technical Field

[0001] This invention relates to an adaptive control method for a digital pump with integrated variable frequency drive, belonging to the field of intelligent control technology. Background Technology

[0002] Digital pumps are intelligent devices that integrate frequency converters, controllers, motors, and pumps. They are compatible with various fluid conveying pipelines and are equipped with multiple media conveying levels to adapt to different conveying flow rates and different media conditions.

[0003] In existing technologies, integrated variable frequency digital pumps generally adopt single-temperature closed-loop speed control logic, relying solely on the temperature of a single medium to adjust the motor speed, without taking the temperature difference between the pump inlet and outlet as an independent control parameter. When the pump delivers a large flow rate for a long time and the medium temperature difference is low, the motor's reactive power loss increases significantly, the equipment's operating efficiency is low, and relying solely on single-path temperature feedback, the speed adjustment response is significantly lagging when the pipeline medium delivery demand changes abruptly, making it impossible to match the actual delivery volume of the pipeline in real time. Summary of the Invention

[0004] To address the shortcomings of existing technologies, the present invention aims to provide an adaptive control method for digital pumps with integrated variable frequency drive. By combining load feedforward prediction with dual PID collaborative regulation, it balances temperature control accuracy and flow load matching, and with speed constraints and steady-state fine-tuning, effectively improving the problems of inefficient operation due to load perception lag and large flow rate with small temperature difference.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] An adaptive control method for a digital pump with integrated variable frequency drive includes:

[0007] It receives the operating gear command and target control parameters, and simultaneously collects the inlet and outlet temperatures and real-time operating frequency of the medium being pumped in the pump.

[0008] Based on the derivation of the optimal temperature difference for heat exchange of the conveying medium, the temperature deviation and flow matching deviation of the conveying medium are calculated respectively.

[0009] Based on the operating gear, inlet and outlet temperatures, and real-time operating frequency, the pump's forward speed is predicted to generate the pump's feedforward reference speed.

[0010] A dual-path PID speed control loop is constructed, and feedback calculations are performed based on temperature deviation and flow matching deviation to generate a corrected reference speed.

[0011] The control speed of the pump is obtained by weighted summation of the feedforward reference speed and the corrected reference speed;

[0012] Based on the pump's maximum allowable speed and minimum guaranteed speed, the control speed is limited to obtain the target speed and then sent to the pump's built-in frequency converter.

[0013] Based on the continuous operation of the pump motor, the inlet and outlet temperature difference of the conveyed medium is updated cyclically. The steady-state cumulative deviation is calculated based on the time-series cache data to make fine-tuning compensation judgment, generate compensation speed and add it to the target speed.

[0014] Specifically, the temperature deviation and flow matching deviation of the conveying medium are calculated separately, including:

[0015] For a single sampling point, the real-time heat exchange temperature difference is obtained by the absolute difference between the inlet and outlet temperatures of the conveying medium;

[0016] Retrieve the target temperature, upper temperature limit, and lower temperature limit of the conveying medium from the target control parameters, and calculate the margin factor based on the proportion of the target temperature within the allowable temperature range;

[0017] The load deviation is obtained by using the absolute difference between the inlet temperature and the target temperature;

[0018] Simultaneously, the rated operating frequency of the digital pump is obtained, and the frequency mapping coefficient is calculated in combination with the real-time operating frequency;

[0019] Based on the margin coefficient and frequency mapping coefficient, the differentiated load control ratio is calculated for the corresponding operating level.

[0020] The minimum effective heat exchange temperature difference of the digital pump is obtained, and the optimal heat exchange temperature difference is obtained by adding the product of the load control ratio and the load deviation to the minimum effective heat exchange temperature difference.

[0021] The temperature deviation is obtained by measuring the difference between the target temperature and the current actual outlet temperature.

[0022] The flow matching deviation is obtained by measuring the difference between the optimal heat exchange temperature difference and the real-time heat exchange temperature difference.

[0023] Specifically, the steps for generating the pump's feedforward reference speed include:

[0024] Based on the sampling sequence within the current period, consecutive inlet temperatures are obtained to construct a short-term load sequence;

[0025] Trend analysis is performed on short-term load sequences to correct the trend of inlet temperature at the latest sampling point;

[0026] Based on the trend-corrected inlet temperature, compare it with the upper and lower temperature limits set by the target user;

[0027] If the inlet temperature is between the upper and lower limits, retain the current trend correction result; otherwise, revert to the original inlet temperature.

[0028] Based on the current operating level, the load difference is calculated using differential subtraction.

[0029] The maximum load adjustment range that the digital pump can withstand in a single cycle is obtained. The load difference is limited by using the maximum load adjustment range as the upper limit and the negative value of the maximum load adjustment range as the lower limit.

[0030] Specifically, the trend correction includes:

[0031] If the inlet temperature in the short-term load sequence shows a continuous upward trend, it is determined that the load is continuously falling, and a positive offset correction is made to the latest sampled inlet temperature.

[0032] If the short-term load sequence shows a continuous downward trend, it is determined that the load is continuously rising, and the latest sampled inlet temperature is corrected by a reverse offset.

[0033] If the inlet temperature in the short-term load sequence does not show a unidirectional rise or fall pattern, it is determined to be a small steady-state disturbance of the load, and the latest inlet temperature is used.

[0034] Specifically, the step of generating the pump's feedforward reference speed also includes:

[0035] Based on the previous closed-loop calculation results of the current cycle, obtain the load difference of the previous week, and combine it with the load difference of this week to calculate the load change rate;

[0036] Obtain the maximum load change rate, and limit the load change rate by setting the maximum load change rate as the upper limit and the negative value of the maximum load change rate as the lower limit.

[0037] The difference in constrained load after slope constraint is calculated in reverse by using the rate of change of the limited load.

[0038] The feedforward correction is obtained by multiplying the constraint load difference with the frequency mapping coefficient, and the gear reference speed is obtained to calculate the feedforward reference speed.

[0039] Specifically, the steps for generating the corrected reference speed include:

[0040] Temperature control PID is set based on temperature deviation, and load PID is set based on flow matching deviation;

[0041] For temperature control PID and load PID, a historical reference sequence is generated by adjusting the historical reference speed to obtain the historical main adjustment direction;

[0042] Based on the real-time control deviation, the expected adjustment direction, including acceleration and deceleration, is determined for temperature control PID and load PID respectively.

[0043] By comparing the expected adjustment direction with the historical main adjustment direction, a retrospective attenuation of the historical adjustment behavior is performed to obtain the control deviation amount after retrospective attenuation; wherein, if the expected adjustment direction is consistent with the historical main adjustment direction, the control deviation amount is assigned an attenuation process; if the expected adjustment direction is opposite to the historical main adjustment direction, the original control deviation amount is used.

[0044] Specifically, the steps for generating the corrected reference speed also include:

[0045] For temperature control PID and load PID, the historical control deviation is obtained and the arithmetic mean is calculated.

[0046] The fluctuation coefficient is obtained by comparing the controlled deviation after backtracking decay with the arithmetic mean.

[0047] Set the lower and upper limits of the volatility coefficient to limit its amplitude.

[0048] The fluctuation coefficient after amplitude limiting is used to make a small correction within the boundary of the proportional term of the PID;

[0049] Continuous monitoring of single-channel PID Whether the output of each control cycle has the same sign; where... This is the default value;

[0050] If all outputs are in phase, the integral term is locked, and no new accumulation is added to the integral term;

[0051] During the locking process of integral terms, when the reverse direction of regulation is first detected, the existing integral amount is gradually released;

[0052] For a PID that has undergone proportional term micro-correction and integral term locking, the control deviation that has been decayed by backtracking is used as input to generate a corrected reference speed, including temperature-corrected speed and load-corrected speed.

[0053] Specifically, the steps for determining the fine-tuning compensation include:

[0054] The total length of the pipeline and the medium delivery velocity in the digital pump are obtained. The time it takes for the medium to flow is obtained by the ratio of the total length to the medium delivery velocity.

[0055] Obtain the control cycle, and then round up the ratio of the medium flow time to the control cycle to get the queue length.

[0056] For each control cycle, the real-time heat exchange temperature difference and the optimal heat exchange temperature difference are integrated, and a cache queue is constructed.

[0057] The arithmetic mean of the real-time heat exchange temperature difference and the optimal heat exchange temperature difference in the buffer queue are calculated to obtain the real-time mean and the theoretical mean, and the steady-state cumulative deviation is calculated.

[0058] Call the rated heat exchange temperature difference of the digital pump and set the instantaneous compensation threshold and cumulative compensation threshold;

[0059] The calculated traffic matching deviation is invoked, and long-term steady-state compensation is performed only if the absolute value of the traffic matching deviation is greater than the instantaneous compensation threshold and the absolute value of the steady-state cumulative deviation is greater than the cumulative compensation threshold.

[0060] Specifically, the long-term steady-state compensation includes:

[0061] Call the steady-state accumulated deviation to determine the compensation direction;

[0062] Obtain the rated speed of the water pump, and compensate for the rated speed by the ratio of steady-state cumulative deviation to rated heat exchange temperature difference to obtain the basic compensation amount, while setting the compensation upper limit.

[0063] Using the aforementioned compensation upper limit, a compensation change upper limit for adjacent periods is set as a gradual adjustment constraint;

[0064] Obtain the effective compensation amount from the previous adjustment cycle, and obtain the adjacent compensation difference by the difference between the basic compensation amount and the effective compensation amount;

[0065] If the absolute value of the difference between adjacent compensations is less than or equal to the upper limit of compensation change, then the basic compensation amount shall be used as the theoretical compensation amount for this regulation cycle.

[0066] If the absolute value of the difference between adjacent compensations is greater than the upper limit of compensation change, then the compensation amount will be adjusted towards the basic compensation amount based on the upper limit of compensation change.

[0067] Specifically, the long-term steady-state compensation also includes:

[0068] A compensation self-locking mechanism is set up to continuously update and monitor the cache queue in order to continuously determine the steady-state cumulative deviation;

[0069] Only when the absolute value of the steady-state cumulative deviation is continuously less than or equal to the cumulative compensation threshold, it is determined that the long-term operating condition deviation has been corrected. In each control cycle, the current effective compensation speed is reduced to zero step by step according to the upper limit of the compensation change, and gradually returned to zero in multiple cycles.

[0070] Once the compensation speed reaches zero, clear the historical compensation and unlock.

[0071] The beneficial effects of this invention are:

[0072] By combining the medium delivery speed setting with the equipment's preset control parameters and the real-time operating conditions of the fluid delivery pipeline, the system pre-calculates the delivery demand, generating a predictive feedforward reference speed. This allows for early adaptation to gradual fluctuations in the pipeline's medium delivery demand, overcoming the limitations of traditional single-mode medium temperature feedback in proactively predicting changes in medium supply and demand. It also avoids the problem of lag in speed control response when medium delivery demand changes abruptly. An independent dual-path PID speed control system is established to specifically correct medium temperature deviation and medium flow matching deviation, simultaneously addressing both precise medium temperature control and pipeline flow matching. This overcomes the limitations of traditional single-loop speed control, which cannot simultaneously balance temperature control accuracy and pump operating efficiency. Furthermore, it eliminates the need for continuous pulling to maintain medium temperature at the feedback calculation level. In situations where high motor speeds result in ineffective media delivery flow, the system ensures the safe and stable operation of the variable frequency motor by hard limiting the pump's rated speed. It utilizes time-series buffer data statistics to achieve fine-tuning compensation for long-term delivery conditions, slightly correcting media delivery deviations caused by long-term pipeline operation. This compensates for the main control loop's insufficient ability to correct steady-state flow mismatches. It adapts to various media delivery conditions, ensuring stable pump temperature control and matching between media supply and demand. It also continuously optimizes the matching degree between pump delivery flow and actual pipeline media requirements, fundamentally suppressing inefficient operation under high flow rates and low media temperature differences. This effectively reduces reactive power losses in the integrated variable frequency digital pump motor and significantly improves the overall operating efficiency and speed control stability of the fluid pipeline under multiple delivery levels. Attached Figure Description

[0073] Figure 1 A flowchart of an adaptive control method for a digital pump with integrated variable frequency drive;

[0074] Figure 2 This is a flowchart illustrating the generation of the feedforward reference rotational speed in this invention;

[0075] Figure 3 This is a flowchart of generating the corrected reference rotational speed in this invention;

[0076] Figure 4 This is a flowchart of the long-term steady-state compensation in this invention. Detailed Implementation

[0077] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present invention and the specific features in the embodiments are detailed descriptions of the technical solution of the present invention, rather than limitations thereof. In the absence of conflict, the embodiments of the present invention and the technical features in the embodiments can be combined with each other.

[0078] Example 1

[0079] refer to Figures 1 to 4 As shown in the figure, this embodiment introduces an adaptive control method for a digital pump with integrated variable frequency drive, including:

[0080] The system obtains the operating mode command and target control parameters issued by the target user through the human-computer interaction interface. The operating mode command includes the operating mode of the digital pump, including heating mode and cooling mode, and the target control parameters include target temperature and upper and lower temperature limits.

[0081] The sensors and frequency converters equipped in the digital pump collect the real-time outlet temperature of the conveyed medium at the pipeline outlet where the digital pump delivers fluid, which is the supply water temperature circulating in the pipeline network; after the fluid completes the pipeline circulation and flows back to the pump input, the real-time inlet temperature of the conveyed medium is collected, which is the return water temperature; at the same time, the real-time operating frequency of the pump set is collected. Combined with the real-time extrapolation mechanism of the optimal heat exchange temperature difference, the real-time heat exchange temperature difference between the supply water and the return water is compared with the optimal heat exchange temperature difference to calculate the control deviation, including temperature deviation and flow matching deviation.

[0082] Based on the operating gear, combined with the return water temperature, target temperature, and real-time operating frequency of the pump set, a forward speed prediction is performed to generate a feedforward reference speed. Two independent PID speed control loops are built to perform feedback calculations on the control deviation and generate a corrected reference speed. By weighted summing the feedforward reference speed and the corrected reference speed, the unconstrained control speed is obtained, achieving the dual control objectives of rapid temperature control and precise flow distribution. Specifically, the temperature control range is obtained by calculating the difference between the upper and lower temperature limits issued by the human-machine interface. The reverse temperature control coefficient is obtained by the ratio of the current absolute value of the temperature deviation to the temperature control range. The temperature control weight of the temperature deviation is obtained by subtracting the reverse temperature control coefficient from 1. Based on the fact that the sum of the weights is 1, the flow matching weight is obtained by the difference between 1 and the temperature control weight.

[0083] The maximum allowable speed and minimum guaranteed speed of the digital pump are obtained, the control speed is limited to obtain the target speed, and the target speed is converted into an analog signal that the frequency converter can recognize. The signal is sent to the frequency converter of the digital pump through the communication interface or analog output channel. The frequency converter analyzes the received analog signal and restores the corresponding drive frequency to drive the digital pump motor to run at the limited target speed.

[0084] During the continuous operation of the digital pump's motor driving the pipeline water circulation, the supply water temperature and return water temperature in the pipeline network are continuously collected. The real-time heat exchange temperature difference between the supply and return water is dynamically updated with the real-time simulation of the optimal heat exchange temperature difference. Based on the time-series cached data, the steady-state cumulative deviation is calculated to make fine-tuning compensation judgment, generate compensation speed, and superimpose the compensation speed into the target speed to complete small-scale dynamic correction. At the same time, data acquisition, deviation calculation, speed calculation and fine-tuning compensation are continuously performed.

[0085] Furthermore, the steps for calculating the control deviation include:

[0086] For a single sampling point, the current real heat exchange temperature difference of the pipeline network is obtained by the absolute difference between the supply water temperature and the return water temperature. This real-time heat exchange temperature difference of the supply and return water is used to characterize the actual heat exchange state of the current water circulation.

[0087] Retrieve the target temperature, upper temperature limit, and lower temperature limit from the target control parameters. Calculate the margin coefficient based on the proportion of the target temperature within the allowable temperature range, i.e., the difference between the target temperature and the lower temperature limit as the numerator and the difference between the lower temperature limit and the upper temperature limit as the denominator, to characterize the temperature operating range position of the current set operating condition.

[0088] The load deviation is obtained by using the absolute difference between the return water temperature and the target temperature, so as to quantify the current required heat / cooling load at the end of the pipe network. The larger the load, the greater the heat exchange temperature difference required by the system to support heat exchange.

[0089] The real-time operating frequency of the pump set is called, and the rated operating frequency of the digital pump is obtained at the same time. The frequency mapping coefficient is obtained by the ratio of the real-time operating frequency of the pump set to the rated operating frequency, which represents the proportion of the current circulating flow to the maximum flow.

[0090] Based on the margin coefficient and frequency mapping coefficient, differentiated load control ratios are calculated for heating and cooling respectively; where the load control ratio for heating is the difference between 1 and the frequency mapping coefficient multiplied by the margin coefficient, and the load control ratio for cooling is the sum of 1 and the frequency mapping coefficient multiplied by the margin coefficient.

[0091] The minimum effective heat exchange temperature difference of the digital pump is obtained by multiplying the load control ratio and the load deviation by the minimum effective heat exchange temperature difference to obtain the optimal heat exchange temperature difference. The minimum effective heat exchange temperature difference is obtained through actual measurement and calibration under normal temperature steady-state no-load conditions of the pump's compatible pipeline. The digital pump is adjusted to the no-load steady-state setting, all branch pipeline media loads are shut off, and the pump is continuously operated at a constant reference speed. Several sets of real-time delivery temperature differences between the pump inlet and outlet are continuously collected at single control cycles, and the maximum and minimum values ​​of the temperature differences are calculated. The maximum and minimum difference is taken as the fluctuation amplitude, and 5% of the rated medium temperature difference is taken as the stability judgment threshold. If the fluctuation amplitude is less than or equal to the stability judgment threshold, all sampled data are considered stable. If the fluctuation amplitude is greater than the stability judgment threshold, the stable operation time is extended, and data is collected again until the fluctuation requirement is met. The arithmetic mean of all stable sample data is taken as the minimum effective delivery temperature difference.

[0092] The temperature deviation is obtained by measuring the difference between the target temperature and the current actual water supply temperature, which is then used to adjust the PID temperature correction.

[0093] The flow matching deviation is obtained by measuring the difference between the optimal heat exchange temperature difference and the real-time heat exchange temperature difference, which helps determine whether the current circulation flow is excessive or insufficient.

[0094] Furthermore, the steps for generating the feedforward reference speed include:

[0095] Since the load on the transported medium is under a continuously varying condition, a single point instantaneous temperature cannot characterize the trend of load change in the pipeline network. Therefore, based on the sampling sequence within the current cycle, continuous temperature data is obtained. Based on the return water temperature, a short-term load sequence is constructed; among which... As a preset value, the one-way length of the medium in the pipeline, after passing through the heat exchange / transport stage in the terminal pipeline, flows back to the pump inlet temperature measuring point along the return pipeline from the farthest measuring point. The one-way return time is obtained by comparing the one-way length with the medium transport velocity. The one-way return time is then rounded up by comparing it with the control cycle to obtain the final value. ;

[0096] Trend analysis is performed on the short-term load sequence to correct the trend of the return water temperature at the latest sampling point. If the return water temperature in the short-term load sequence shows a continuous upward trend (i.e., the slope of the overall sequence fitting is positive), it is determined that the terminal load is continuously declining, and a positive offset correction is made to the latest sampled return water temperature. This involves adding an offset to the return water temperature to moderately increase the equivalent return water temperature value and weaken the calculated heating and cooling load demand. Conversely, if the short-term load sequence shows a continuous downward trend (i.e., the slope of the overall sequence fitting is negative), it is determined that the terminal load is continuously rising, and a negative offset correction is made to the latest sampled return water temperature. The calculation involves subtracting the offset from the return water temperature, appropriately lowering the equivalent return water temperature value, and amplifying the calculated heating and cooling load demand. If the return water temperature in the short-term load sequence does not have a fixed unidirectional rise and fall pattern, it is determined to be a small-scale steady-state disturbance of the pipeline load, and no temperature offset correction operation is performed; the latest original return water temperature is directly used. In this process, the linkage correction coefficient matching the hysteresis characteristics of the pipeline is obtained, which is the ratio of the single-pass return time to the medium flow time. At the same time, the sequence slope of the short-term load sequence is fitted by the least squares method, and the offset is obtained by multiplying the linkage correction coefficient and the absolute value of the sequence slope.

[0097] Based on the corrected return water temperature, compare it with the upper and lower temperature limits set by the target user; if the return water temperature is between the upper and lower temperature limits, it indicates that the value after this trend correction is within the reasonable range of the hardware temperature measurement range and the system process, and the parameter has physical and engineering significance, so the result of this trend correction is retained; if the return water temperature is not between the upper and lower temperature limits, it indicates that the trend offset calculation has generated an invalid value that exceeds the range, so this trend correction is invalidated, and a forced rollback is performed to use the original return water temperature.

[0098] Based on the current operating level, the load difference is calculated using differential subtraction. For heating mode, the load difference is the difference between the target temperature and the corrected return water temperature. A positive difference indicates that the return water temperature is too low, resulting in a heat load shortfall, requiring an increase in speed. A negative difference indicates that the return water temperature is too high, resulting in excess heat, requiring a decrease in speed. For cooling mode, the load difference is the difference between the corrected return water temperature and the target temperature. A positive difference indicates that the return water temperature is too high, resulting in a cooling load shortfall, requiring an increase in speed. A negative difference indicates that the return water temperature is too low, resulting in excess cooling capacity, requiring a decrease in speed.

[0099] By conducting full-condition testing of the target-compatible pipeline fluid, the maximum load adjustment range that the digital pump can withstand in a single cycle is obtained. The load difference is then limited, with the maximum load adjustment range as the upper limit and the negative of the maximum load adjustment range as the lower limit. The limiting process includes: if the load difference is greater than the upper limit, the upper limit is forcibly assigned to the load difference, i.e., the load difference is set to the upper limit; if the load difference is less than the lower limit, the lower limit is forcibly assigned to the load difference; if the load difference is between the lower and upper limits, the original load difference is maintained.

[0100] Based on the previous closed-loop calculation results of the current cycle, obtain the load difference of the previous week's period and combine it with the load difference of this week's period to calculate the load change rate, which is the difference between the load difference of this week's period and the load difference of the previous week's period divided by the control cycle.

[0101] Obtain the maximum load change rate calibrated at the factory of the digital pump. With the maximum load change rate as the upper limit and the negative of the maximum load change rate as the lower limit, the load change rate is limited. The load change rate is then used to calculate the constraint load difference after slope constraint in reverse by the limited load change rate. That is, the product of the upper limit load difference plus the load change rate after the limit and the control cycle.

[0102] By using the proportional coefficient between temperature difference and rotational speed, the constraint load difference is mapped to rotational speed to obtain the reference offset. The product of the reference offset and the frequency mapping coefficient is used to obtain the feedforward correction amount. At the same time, the gear reference speed is obtained. The feedforward reference speed is obtained by summing the feedforward correction amount and the gear reference speed. In this process, several sets of temperature difference gradients are set through factory no-load tests, and the rotational speed change under each temperature difference is recorded. The proportional coefficient is obtained by the ratio of the rotational speed change to the temperature difference change. The gear reference speed is calibrated by actual measurement during pipeline no-load operation.

[0103] Furthermore, the steps for generating the corrected reference speed include:

[0104] Two incremental PID calculation units with no data interaction and isolated parameter storage spaces are divided. The first PID loop is bound to the temperature deviation as the only input, i.e., temperature control PID, to eliminate the deviation between the supply water temperature and the user-set target temperature and ensure basic temperature control accuracy. The second PID loop is bound to the flow matching deviation as the only input, i.e., load PID, to reduce the difference between the theoretical optimal temperature difference and the actual supply and return water temperature difference, and to suppress the inefficient water delivery condition of large flow and small temperature difference from the control logic. For both PIDs, a physical test bench is built, the derivative and integral actions are disconnected, and only the proportional part is kept to run in a closed loop. The proportional coefficient is gradually increased from small to large, and the other parameters are set to zero. The supply and return water temperature and pump speed curves are continuously observed until a constant amplitude continuous oscillation occurs. The critical proportional gain and the single oscillation period are recorded at this time. According to the Ziegler-Nichols standard empirical formula, the initial parameters of the PID are directly calculated. That is, the temperature control PID performs the above step oscillation test on the supply water temperature-set temperature closed loop alone, and the load PID performs the test on the actual temperature difference-theoretical optimal temperature difference closed loop alone.

[0105] For temperature control PID and load PID, the adjustment direction of the most recent 3 control cycles is obtained to generate a historical reference sequence. The direction with the most proportion in the sequence is taken as the historical main adjustment direction. The control cycle is the time to complete the entire process of data acquisition, deviation calculation, feedforward calculation, PID calculation and speed output. The adjustment direction is the sign of the correction reference speed generated by the single PID in the previous control cycle. If it is positive, the adjustment direction is to increase speed. If it is negative, the adjustment direction is to decrease speed.

[0106] Based on the control deviation, the expected adjustment direction is determined for both the temperature control PID and the load PID. When the temperature deviation is positive, it indicates that the water supply temperature of the temperature control PID is low, requiring an increase in pump speed and circulation flow for heat replenishment; the expected adjustment direction is to increase speed. When the temperature deviation is negative, the water supply temperature is higher than the set value, requiring the system to reduce speed and flow; the expected adjustment direction is to decrease speed. When the flow matching deviation is positive, it indicates that the actual temperature difference is less than the optimal heat exchange temperature difference, resulting in excessive flow and redundancy; the expected adjustment direction is to decrease speed. When the flow matching deviation is negative, it indicates that the actual temperature difference is greater than the theoretical optimal temperature difference, resulting in insufficient flow and inadequate heat exchange; the expected adjustment direction is to increase speed.

[0107] By comparing the current expected adjustment direction with the historical main adjustment direction, and through scenario determination, the historical adjustment behavior is retrospectively attenuated to obtain the control deviation amount after retrospective attenuation. Specifically, if the expected adjustment direction is consistent with the historical main adjustment direction, it indicates that the loop has continuously adjusted the speed in the same direction for multiple cycles. The thermal inertia of the pipeline network will cause the adjustment effect to be superimposed, which is very easy to cause over-adjustment, high speed, and redundant flow. The control deviation amount is assigned an attenuation processing. If the expected adjustment direction is opposite to the historical main adjustment direction, it indicates that the deviation has begun to converge in the opposite direction and the over-adjustment amount is about to be corrected. There is no need to constrain the adjustment force. At this time, no attenuation is performed, and the original control deviation amount is directly used.

[0108] For temperature control PID and load PID, the control deviation of the corresponding loop for the last 5 control cycles is obtained and the arithmetic mean is calculated. The fluctuation coefficient is obtained by the ratio of the control deviation after backtracking decay to the arithmetic mean.

[0109] To avoid oscillations caused by drastic parameter changes, the fine-tuning range of the proportional term is limited to no more than ±20% of the original proportional coefficient. Therefore, 0.8 and 1.2 are used as the lower and upper limits of the fluctuation coefficient, respectively, to limit the fluctuation coefficient. By multiplying the fluctuation coefficient after the limit treatment with the original proportional coefficient, a small correction is made to the proportional term of the PID within the boundary.

[0110] Continuous monitoring of continuous single-channel PID If the outputs of each control cycle are of the same sign, and all are in-phase outputs, then the integral is deemed to have a risk of saturation accumulation. In this case, the integral term is locked, and the current accumulated value is locked. No new accumulated value is added to the integral term in this cycle. This is a preset value, obtained by rounding down the ratio of the medium flow time to the control cycle;

[0111] During the integral term locking process, when the control direction is detected to be reversed for the first time, the integral is not directly cleared to zero. Instead, a preset integral deduction is deducted for each control cycle to gradually release the integral inventory and slowly unfreeze it. Specifically, the total amount of integral accumulation inside the integral register when the target speed reaches the pump's maximum allowable speed is obtained as the maximum integral accumulation value. The integral deduction is obtained by the ratio of the maximum integral accumulation value to the queue length.

[0112] For a PID that has undergone proportional term micro-correction and integral term locking, the control deviation that has been decayed by backtracking is used as input to generate a corrected reference speed, including temperature-corrected speed and load-corrected speed.

[0113] Furthermore, the steps for fine-tuning the compensation determination include:

[0114] The total length of the pipeline and the medium delivery velocity in the digital pump are obtained. The time it takes for the medium to flow is obtained by the ratio of the total length to the medium delivery velocity.

[0115] The control cycle is obtained by comparing the medium flow time with the control cycle and rounding the comparison value up to obtain the queue length. This ensures that the operating data of all control cycles within the time range of the medium completing a whole closed-loop pipeline journey are covered, so that the cached sample can contain all heat exchange lag feedback information at the end. It will not lose lag conditions due to the queue being too short, nor will it include expired and invalid historical data due to the queue being too long.

[0116] For each control cycle, the real-time heat exchange temperature difference and the optimal heat exchange temperature difference of the supply and return water are integrated, and a buffer queue is constructed. If the length of the buffer queue has reached the queue length, the data at the head of the buffer queue is automatically removed, and the corresponding data is written to the tail of the queue.

[0117] The arithmetic mean of the real-time heat exchange temperature difference and the optimal heat exchange temperature difference in the buffer queue are calculated to obtain the real-time mean and the theoretical mean. The steady-state cumulative deviation is obtained by the difference between the theoretical mean and the real-time mean.

[0118] The rated heat exchange temperature difference of the digital pump is used, and a single-point deviation within 15% of the rated heat exchange temperature difference is taken as normal random fluctuation. An instantaneous compensation threshold is set. Once the flow matching deviation is greater than the instantaneous compensation threshold, it indicates that the single disturbance exceeds the normal noise range.

[0119] The average deviation exceeding the rated heat exchange temperature difference by 25% within a complete pipeline lag cycle is used as the correction boundary. An accumulated compensation threshold is set. Once the steady-state accumulated deviation exceeds the accumulated compensation threshold, it indicates that a continuous flow mismatch has been formed. Long-term operation will inevitably result in considerable energy waste, and fine-tuning correction is necessary.

[0120] The calculated flow matching deviation is called, and the absolute value of the flow matching deviation is used for preliminary screening of the instantaneous deviation. If the absolute value of the flow matching deviation is less than or equal to the instantaneous compensation threshold, it is determined that there is only a normal single-point random disturbance. All subsequent steady-state judgment and compensation processes are skipped directly, and no fine-tuning is performed in this control cycle. The target speed is directly used.

[0121] If the absolute value of the flow matching deviation is greater than the instantaneous compensation threshold, it means that the single instantaneous disturbance exceeds the reasonable noise range, and the steady-state cumulative deviation judgment is entered. If the absolute value of the steady-state cumulative deviation is less than or equal to the cumulative compensation threshold, it indicates that there is only a single abnormal disturbance and no long-term steady-state flow mismatch. The disturbance record is marked, the compensation speed is not calculated, the fine adjustment is not added, and the target speed is kept unchanged. If the absolute value of the steady-state cumulative deviation is greater than the cumulative compensation threshold, after a whole period of return water lag, it is confirmed that there is a continuous flow redundancy or flow loss. The compensation speed calculation and iterative superposition process is formally entered, and long-term steady-state compensation is performed.

[0122] Furthermore, the steps of long-term steady-state compensation include:

[0123] Call the steady-state cumulative deviation to determine the compensation direction; if the steady-state cumulative deviation is positive, it indicates that the average real-time heat exchange temperature difference in the sequence is lower than the average optimal heat exchange temperature difference for a long time, the overall circulation flow is too large, and it is in a low-efficiency condition of large flow and small temperature difference. The compensation direction is to reduce the speed; if the steady-state cumulative deviation is negative, it indicates that the average real-time heat exchange temperature difference in the sequence is higher than the average optimal heat exchange temperature difference for a long time, the circulation flow is too small, and the end heat exchange is insufficient. The compensation direction is to increase the speed.

[0124] The rated speed of the water pump is obtained. The rated speed is compensated by the ratio of the steady-state cumulative deviation to the rated heat exchange temperature difference to obtain the basic compensation amount, which is the ratio of the steady-state cumulative deviation to the rated heat exchange temperature difference multiplied by the rated speed. At the same time, the single fine adjustment is limited to no more than 8% of the rated speed to prevent the fine adjustment level from covering the upstream feedforward and PID coarse adjustment results, and to ensure that the main control priority is higher than the subsequent steady-state fine adjustment, thereby obtaining the compensation upper limit.

[0125] Since directly using the new compensation amount in this control cycle would cause a step jump in the compensation amount, the pipeline flow would change instantaneously, resulting in water hammer impact and instantaneous flow redundancy. Therefore, the change step size of the compensation amount in adjacent cycles is limited. A multi-cycle gradual approximation of the target compensation value is adopted, that is, the constraint is carried out in cycle iterations. 0.4 times the compensation upper limit is used as the upper limit of the compensation change in adjacent cycles as a gradual adjustment constraint.

[0126] The compensation speed that is finally superimposed on the target speed in the previous control cycle, sent to the frequency converter and actually driven the motor to run is recorded as the effective compensation amount. The difference between the basic compensation amount and the effective compensation amount is calculated as the adjacent compensation difference.

[0127] If the absolute value of the difference between adjacent compensations is less than or equal to the upper limit of compensation change, then the basic compensation amount shall be used as the theoretical compensation amount for this regulation cycle.

[0128] If the absolute value of the adjacent compensation difference is greater than the upper limit of compensation change, then the compensation amount will be adjusted towards the basic compensation amount based on the upper limit of compensation change. That is, when the adjacent compensation difference is positive, the theoretical compensation amount is the sum of the effective compensation amount and the upper limit of compensation change; when the adjacent compensation difference is negative, the theoretical compensation amount is the difference between the effective compensation amount and the upper limit of compensation change.

[0129] During this control cycle, fine-tuning instructions are issued based on the theoretical compensation amount, and the theoretical compensation amount is used as a new effective compensation amount for constraint to avoid a sudden jump to the target compensation value and abrupt instruction changes.

[0130] A compensation self-locking mechanism is set up for periodic iterations to continuously update and monitor the cache queue in order to continuously determine the steady-state cumulative deviation;

[0131] Only when the absolute value of the steady-state cumulative deviation is continuously less than or equal to the cumulative compensation threshold, it is determined that the long-term operating condition deviation has been completely corrected. The steady-state compensation needs to be gradually withdrawn, and the currently effective compensation speed should be reduced to zero in each control cycle according to the compensation change upper limit as the step size, and gradually returned to zero in multiple cycles.

[0132] Once the compensation speed reaches zero, the historical compensation is cleared and the system is unlocked. If steady-state compensation is triggered again, the iterative compensation process will be restarted.

[0133] Working principle and effects:

[0134] By issuing multiple media delivery levels and equipment control parameters via human-machine interface, the system simultaneously collects pump outlet medium temperature, pump inlet medium temperature, and real-time pump operating frequency. Based on the optimal delivery temperature difference, it calculates and precisely adjusts deviations in real time. It generates a feedforward reference speed with predictive capabilities based on pre-calculation of media delivery demand, enabling early response to fluctuations in pipeline media supply and demand. This effectively solves the problem of delayed perception of delivery demand in traditional single-temperature speed regulation. Simultaneously, it establishes two independent PID speed regulation systems to correct media temperature deviation and media flow matching deviation, simultaneously considering both media temperature control accuracy and delivery flow matching requirements. From the underlying speed regulation logic, it avoids the generation of excessive and ineffective delivery flow, suppresses inefficient operation under high flow and low media temperature difference conditions, and ensures stable motor operation in conjunction with upper and lower limits of pump rated speed constraints. Furthermore, it performs long-term fine-tuning compensation based on time-series cached operating data to correct media delivery deviations formed during long-term pipeline operation. This adapts to various media delivery levels and operating conditions, significantly improving the matching degree between pump delivery flow and media demand, reducing the reactive power consumption of the integrated variable frequency digital pump motor, and overall improving the energy efficiency and speed regulation stability of the pump equipment under all operating conditions.

[0135] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. An adaptive control method for a digital pump with integrated variable frequency drive, characterized in that, include: It receives the operating gear command and target control parameters, and simultaneously collects the inlet and outlet temperatures and real-time operating frequency of the medium being pumped in the pump. Based on the derivation of the optimal temperature difference for heat exchange of the conveying medium, the temperature deviation and flow matching deviation of the conveying medium are calculated respectively. Based on the operating gear, inlet and outlet temperatures, and real-time operating frequency, the pump's forward speed is predicted to generate the pump's feedforward reference speed. A dual-path PID speed control loop is constructed, and feedback calculations are performed based on temperature deviation and flow matching deviation to generate a corrected reference speed. The control speed of the pump is obtained by weighted summation of the feedforward reference speed and the corrected reference speed; Based on the pump's maximum allowable speed and minimum guaranteed speed, the control speed is limited to obtain the target speed and then sent to the pump's built-in frequency converter. Based on the continuous operation of the pump motor, the inlet and outlet temperature difference of the conveyed medium is updated cyclically. The steady-state cumulative deviation is calculated based on the time-series cache data to make fine-tuning compensation judgments, generate compensation speeds and add them to the target speed.

2. The adaptive control method for a digital pump with integrated variable frequency drive according to claim 1, characterized in that, Calculate the temperature deviation and flow matching deviation of the conveying medium separately, including: For a single sampling point, the real-time heat exchange temperature difference is obtained by the absolute difference between the inlet and outlet temperatures of the conveying medium; Retrieve the target temperature, upper temperature limit, and lower temperature limit of the conveying medium from the target control parameters, and calculate the margin factor based on the proportion of the target temperature within the allowable temperature range; The load deviation is obtained by using the absolute difference between the inlet temperature and the target temperature; Simultaneously, the rated operating frequency of the digital pump is obtained, and the frequency mapping coefficient is calculated in combination with the real-time operating frequency; Based on the margin coefficient and frequency mapping coefficient, the differentiated load control ratio is calculated for the corresponding operating level. The minimum effective heat exchange temperature difference of the digital pump is obtained, and the optimal heat exchange temperature difference is obtained by adding the product of the load control ratio and the load deviation to the minimum effective heat exchange temperature difference. The temperature deviation is obtained by measuring the difference between the target temperature and the current actual outlet temperature. The flow matching deviation is obtained by measuring the difference between the optimal heat exchange temperature difference and the real-time heat exchange temperature difference.

3. The adaptive control method for a digital pump with integrated variable frequency drive according to claim 2, characterized in that, The steps for generating the pump's feedforward reference speed include: Based on the sampling order within the current period, obtain continuous Based on the inlet temperature, a short-term load sequence is constructed; Trend analysis is performed on short-term load sequences to correct the trend of inlet temperature at the latest sampling point; Based on the trend-corrected inlet temperature, compare it with the upper and lower temperature limits set by the target user; If the inlet temperature is between the upper and lower limits, retain the current trend correction result; otherwise, revert to the original inlet temperature. Based on the current operating level, the load difference is calculated using differential subtraction. The maximum load adjustment range that the digital pump can withstand in a single cycle is obtained. The load difference is limited by using the maximum load adjustment range as the upper limit and the negative value of the maximum load adjustment range as the lower limit.

4. The adaptive control method for a digital pump with integrated variable frequency drive according to claim 3, characterized in that, The trend correction includes: If the inlet temperature in the short-term load sequence shows a continuous upward trend, it is determined that the load is continuously falling, and a positive offset correction is made to the latest sampled inlet temperature. If the short-term load sequence shows a continuous downward trend, it is determined that the load is continuously rising, and the latest sampled inlet temperature is corrected by a reverse offset. If the inlet temperature in the short-term load sequence does not show a unidirectional rise or fall pattern, it is determined to be a small steady-state disturbance of the load, and the latest inlet temperature is used.

5. The adaptive control method for a digital pump with integrated variable frequency drive according to claim 4, characterized in that, The steps for generating the pump's feedforward reference speed also include: Based on the previous closed-loop calculation results of the current cycle, obtain the load difference of the previous week, and combine it with the load difference of this week to calculate the load change rate; Obtain the maximum load change rate, and limit the load change rate by setting the maximum load change rate as the upper limit and the negative value of the maximum load change rate as the lower limit. The difference in constrained load after slope constraint is calculated in reverse by using the rate of change of the limited load. The feedforward correction is obtained by multiplying the constraint load difference with the frequency mapping coefficient, and the gear reference speed is obtained to calculate the feedforward reference speed.

6. The adaptive control method for a digital pump with integrated variable frequency drive according to claim 5, characterized in that, The steps for generating the corrected reference speed include: Temperature control PID is set based on temperature deviation, and load PID is set based on flow matching deviation; For temperature control PID and load PID, a historical reference sequence is generated by adjusting the historical reference speed to obtain the historical main adjustment direction; Based on the real-time control deviation, the expected adjustment direction, including acceleration and deceleration, is determined for temperature control PID and load PID respectively. By comparing the expected adjustment direction with the historical main adjustment direction, a retrospective attenuation of the historical adjustment behavior is performed to obtain the control deviation amount after retrospective attenuation; wherein, if the expected adjustment direction is consistent with the historical main adjustment direction, the control deviation amount is assigned an attenuation processing; if the expected adjustment direction is opposite to the historical main adjustment direction, the original control deviation amount is used.

7. The adaptive control method for a digital pump with integrated variable frequency drive according to claim 6, characterized in that, The steps for generating the corrected reference speed also include: For temperature control PID and load PID, the historical control deviation is obtained and the arithmetic mean is calculated. The fluctuation coefficient is obtained by comparing the controlled deviation after backtracking decay with the arithmetic mean. Set the lower and upper limits of the volatility coefficient to limit its amplitude. The fluctuation coefficient after amplitude limiting is used to make a small correction within the boundary of the proportional term of the PID; Continuous monitoring of continuous single-channel PID Whether the output of each control cycle has the same sign; where... This is the default value; If all outputs are in phase, the integral term is locked, and no new accumulation is added to the integral term; During the locking process of integral terms, when the reverse direction of regulation is first detected, the existing integral amount is gradually released; For a PID that has undergone proportional term micro-correction and integral term locking, the control deviation that has been decayed by backtracking is used as input to generate a corrected reference speed, including temperature-corrected speed and load-corrected speed.

8. The adaptive control method for a digital pump with integrated variable frequency drive according to claim 7, characterized in that, The steps for determining the fine-tuning compensation include: The total length of the pipeline and the medium delivery velocity in the digital pump are obtained. The time it takes for the medium to flow is obtained by the ratio of the total length to the medium delivery velocity. Obtain the control cycle, and then round up the ratio of the medium flow time to the control cycle to get the queue length. For each control cycle, the real-time heat exchange temperature difference and the optimal heat exchange temperature difference are integrated, and a cache queue is constructed. The arithmetic mean of the real-time heat exchange temperature difference and the optimal heat exchange temperature difference in the buffer queue are calculated to obtain the real-time mean and the theoretical mean, and the steady-state cumulative deviation is calculated. Call the rated heat exchange temperature difference of the digital pump and set the instantaneous compensation threshold and cumulative compensation threshold; The calculated traffic matching deviation is invoked, and long-term steady-state compensation is performed only if the absolute value of the traffic matching deviation is greater than the instantaneous compensation threshold and the absolute value of the steady-state cumulative deviation is greater than the cumulative compensation threshold.

9. The adaptive control method for a digital pump with integrated variable frequency drive according to claim 8, characterized in that, The long-term steady-state compensation includes: Call the steady-state accumulated deviation to determine the compensation direction; Obtain the rated speed of the water pump, and compensate for the rated speed by the ratio of steady-state cumulative deviation to rated heat exchange temperature difference to obtain the basic compensation amount, while setting the compensation upper limit. Using the aforementioned compensation upper limit, a compensation change upper limit for adjacent periods is set as a gradual adjustment constraint; Obtain the effective compensation amount from the previous adjustment cycle, and obtain the adjacent compensation difference by the difference between the basic compensation amount and the effective compensation amount; If the absolute value of the difference between adjacent compensations is less than or equal to the upper limit of compensation change, then the basic compensation amount shall be used as the theoretical compensation amount for this regulation cycle. If the absolute value of the difference between adjacent compensations is greater than the upper limit of compensation change, then the compensation amount will be adjusted towards the basic compensation amount based on the upper limit of compensation change.

10. The adaptive control method for a digital pump with integrated variable frequency drive according to claim 9, characterized in that, The long-term steady-state compensation also includes: A compensation self-locking mechanism is set up to continuously update and monitor the cache queue in order to continuously determine the steady-state cumulative deviation; Only when the absolute value of the steady-state cumulative deviation is continuously less than or equal to the cumulative compensation threshold, it is determined that the long-term operating condition deviation has been corrected. In each control cycle, the current effective compensation speed is reduced to zero step by step according to the upper limit of the compensation change, and gradually returned to zero in multiple cycles. Once the compensation speed reaches zero, clear the historical compensation and unlock.

Images