Filter performance optimization system based on cfau multi-module

Abstract

The application relates to the technical field of air filtration and intelligent control, and discloses a filtering performance optimization system based on a CFAU multi-module, which comprises four modules of C-Unit acquisition, F-Unit feedback, A-Unit decision and U-Unit execution. The C-Unit acquires dust concentrations in front of and behind a filter element in real time and adaptively compresses data, and synchronously acquires current. The F-Unit periodically constructs a differential pressure grid and calculates filter element load, instantaneous filtering efficiency and unit air volume energy consumption. The A-Unit generates a candidate scheme according to a double-threshold model of efficiency-energy consumption, issues an instruction when the double threshold is met, and reduces the gear otherwise. The U-Unit returns an energy consumption margin after adjusting the rotating speed of a fan, an air duct or electrostatic field intensity, and closes the cycle. The system can guarantee filtering efficiency, lower energy consumption and prolong the service life of the filter element, and can be used for intelligent upgrading of air purification equipment.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of air filtration and intelligent control, in particular to a filtration performance optimization system based on CFAU multi-module. BACKGROUND

[0002] The prototype of air filtration and purification technology can be traced back to the period when coal dust hazards were widely concerned at the end of the 19th century. At that time, simple filter cloth made of cotton fiber or woolen fabric began to appear in industrial sites to block visible particles. In the 20th century, with the gradual maturity of high molecular fiber, glass fiber and electrostatic polarization technology, various types of high-efficiency particle air filters (HEPA, ULPA) were commercialized one after another, and the characterization of filtration efficiency also evolved from the early "weight capture rate" to the number of particles counted in different particle size ranges. After the 1990s, the frequency conversion of fans, the brushless motor and the cost reduction of microcontrollers made it possible to add a basic air volume-differential pressure closed loop to the filtration equipment. Some manufacturers began to introduce sensors to upload data such as dust concentration, pressure difference and current curve at the inlet and back end to the control board to achieve "automatic transmission" and "filter core blockage alarm" with coarse-grained logic. In the era of Internet of Things and cloud computing, more devices push data to the cloud through 4G / Wi-Fi, and use mobile applications to view air quality, filter core life and energy consumption reports; at the same time, the breakthrough of deep learning in image and voice fields has prompted academia and industry to try to use convolutional neural networks and time series networks to predict filter core blockage trends or optimal replacement time.

[0003] However, existing air filtration equipment generally exposes two deficiencies in long-term operation: first, there is a lack of flexible compression and high-frequency processing mechanism for multi-particle size particle concentration data. The traditional approach either uploads all counts of the six particle size channels with high bandwidth and MCU computing power consumption, or only uploads the total concentration, which makes it impossible to detect the sudden increase of fine particle proportion and the penetration risk. Second, there is a lack of rapid decision-making logic that takes into account the lower limit of filtration efficiency and the upper limit of energy consumption. The current "high-speed gear" and "silent gear" switching of most devices only relies on single threshold triggering, ignoring the real-time coupling relationship between filter core partial blockage and energy consumption margin, resulting in either efficiency meeting standards but high energy consumption, or energy consumption controlled but unable to maintain cleanliness. Especially when the filter core is gradually saturated or the external particle distribution mutates, this single-index-driven strategy is prone to excessive gear shifting or delayed response, shortening the filter core life and increasing the use cost. Therefore, a filtration performance optimization system is needed that can dynamically reduce data and power consumption while ensuring multi-particle size filtration efficiency. SUMMARY

[0004] (I) The technical problem solved: In view of the deficiencies of the prior art, the present application provides a filtration performance optimization system based on CFAU multi-module, which solves the above problems.

[0005] (ii) Technical solution: To achieve the above object, the present application provides the following technical solution: a filtering performance optimization system based on CFAU multi-module, the system comprises: a C-Unit acquisition module, used for acquiring dust concentration data at the front end and rear end of the filter element, synchronously acquiring current data of the power supply line, and stamping a unified time stamp on the dust concentration data and the current data; performing adaptive interval processing on the dust concentration data collected in a preset time window at the front end of the filter element, to dynamically adjust the number of particle size intervals at the front end and rear end of the filter element and the particle size count vector according to the dust concentration fluctuation degree.

[0006] A F-Unit feedback evaluation module: used for acquiring multi-point differential pressure at the rear end of the filter element in a preset evaluation period and forming differential pressure grid data, and generating a filter element partition load matrix, instantaneous filtration efficiency and unit air volume energy consumption based on the dust concentration data, current data and differential pressure grid data.

[0007] An A-Unit adaptive decision module: used for generating a candidate control scheme according to the system preset lower limit of filtration efficiency and upper limit of unit air volume energy consumption; adopting a double-threshold negotiation mechanism to output an execution instruction only when the candidate scheme meets the filtration efficiency and energy consumption constraints at the same time, otherwise outputting a downshift instruction.

[0008] A U-Unit execution update module: used for receiving the execution instruction to adjust the fan speed, switch the air duct or adjust the electrostatic field strength, and feeding back the energy consumption margin to the A-Unit adaptive decision module after execution.

[0009] Preferably, the C-Unit acquisition module comprises: a dust concentration acquisition unit, used for acquiring dust concentration data at the front end and rear end of the filter element at a preset sampling frequency.

[0010] A current detection unit, used for synchronously acquiring current data of the power supply line.

[0011] A time synchronization unit, used for time stamping the dust concentration data and the current data.

[0012] An adaptive interval unit, used for statistically analyzing the dust concentration data in a preset time window at the front end of the filter element, dynamically adjusting the number of particle size intervals at the front end and rear end of the filter element according to the dust concentration fluctuation degree, and reorganizing the particle size count vector according to the adjusted number of particle size intervals.

[0013] Preferably, the adaptive interval unit comprises: a variance calculation subunit, used for calculating the variance of the dust concentration data in a preset time window at the front end of the filter element.

[0014] The threshold comparison subunit is configured to compare the variance with a preset variance threshold, and if the variance is less than the preset variance threshold, determine that the number of particle size intervals is halved, and if the variance is greater than or equal to the threshold, keep the number of particle size intervals unchanged.

[0015] The interval merging subunit is configured to, when it is determined that the number of particle size intervals is halved, merge adjacent intervals in pairs in sequence and add the count values of the adjacent two particle size intervals to obtain a reorganized particle size count vector at the front end and the rear end of the filter element.

[0016] Preferably, the F-Unit feedback evaluation module comprises a differential pressure grid acquisition unit configured to collect multi-point differential pressures at the rear end of the filter element according to a preset sampling period and form differential pressure grid data.

[0017] The load matrix calculation unit is configured to compare the differential pressure grid data with a reference differential pressure of the filter element to generate a filter element partition load matrix.

[0018] The filtration efficiency calculation unit is configured to calculate an instantaneous filtration efficiency according to the particle size count vectors at the front end and the rear end of the reorganized filter element.

[0019] The energy consumption calculation unit is configured to calculate a unit air volume energy consumption according to the supply line current data and an average differential pressure value of the differential pressure grid data.

[0020] Preferably, the A-Unit adaptive decision module comprises a control scheme generation unit configured to combine at least one of a fan gear, an air duct channel and an electrostatic field intensity according to a preset control rule to form a candidate control scheme, and estimate an efficiency improvement amount and an energy consumption increase amount corresponding to the candidate control scheme.

[0021] The double threshold comparison unit is configured to compare the efficiency improvement amount with a lower limit of system filtration efficiency, and compare the energy consumption increase amount with an upper limit of current unit air volume energy consumption.

[0022] The instruction output unit is configured to output a candidate control scheme execution instruction when both thresholds are satisfied, and output a gear down instruction when either threshold is not satisfied.

[0023] Preferably, the preset control rule comprises: when the number of high load blocks in the filter element partition load matrix exceeds a first preset threshold, preferentially generating a candidate control scheme of switching the air duct channel.

[0024] When the instantaneous filtration efficiency gap is less than a preset efficiency compensation threshold and the energy consumption margin is greater than a second preset threshold, preferentially generating a control scheme of improving the electrostatic field intensity.

[0025] When none of the above conditions is satisfied, generating a control scheme of improving the fan gear.

[0026] Preferably, the control scheme generation unit calls an efficiency-energy consumption approximate model based on experimental calibration by look-up table to estimate the efficiency improvement and energy consumption increase of each candidate scheme.

[0027] Preferably, the double threshold comparison unit starts a negotiation time window after outputting the downshift instruction, and re-triggers the control scheme generation unit to generate a new control scheme candidate and repeats the comparison process of the double threshold comparison unit when the negotiation time window expires.

[0028] Preferably, the double threshold comparison unit compares by the following relationship: ; wherein, is the lower limit of system filtration efficiency, is the instantaneous filtration efficiency, is the efficiency improvement of the candidate control scheme, is the current unit wind energy consumption, is the energy consumption increase of the candidate control scheme, is the upper limit of the current unit wind energy consumption.

[0029] (Three) beneficial effects: compared with the prior art, the present application provides a filtration performance optimization system based on CFAU multi-module, which has the following beneficial effects: 1. The filtration performance optimization system based on CFAU multi-module, by proposing an "adaptive interval processing" mechanism in the C-Unit acquisition module, the dust concentration sequence in the continuous time window at the front end of the filter element is subjected to variance detection, the air quality fluctuation level is automatically determined, and the number of particle size intervals is dynamically combined or maintained accordingly, when the environment is stable, adjacent particle size buckets are combined into half the number of intervals according to the rules, and when the environment suddenly changes, all intervals are immediately restored, thereby retaining the resolution of fine particle distribution, and reducing the uplink data volume and MCU operation volume by about half in a stable scenario. Compared with the prior art, the present application either uploads all dust concentration data causing communication and storage redundancy, or only uploads the total concentration and loses the details of the particle size, the present application realizes real-time adaptive balance of particle size information and resource consumption between different working conditions; the lengths of the front and rear particle size vectors are always consistent, so that the subsequent filtration efficiency calculation does not need to be interpolated or supplemented twice, significantly improving the real-time performance and reliability of the local side.

[0030] 2、The filtering performance optimization system based on the CFAU multi-module, by constructing the "efficiency-energy consumption double threshold negotiation" control logic in the A-Unit adaptive decision module, first, the F-Unit gives the real-time filtering efficiency and filter load, then the U-Unit feedbacks the current energy consumption margin, then the efficiency improvement of the candidate control scheme is compared with the lower limit of the system filtering efficiency, and the energy consumption increase is compared with the upper limit of the unit air volume energy consumption, only when both meet, the execution instruction is output, otherwise, the downshift is entered and the negotiation timing window is started. This method avoids the "efficiency enough but power consumption out of control" or "low power consumption but poor filtering" problem caused by traditional single threshold triggering, and can still realize self-correction in the scene of gradual clogging of the filter element or sudden change of particle distribution, ensuring that the filtering efficiency continues to meet the standard while the energy consumption peak is suppressed within the design threshold; at the same time, the mechanical impact of frequent high-low shift on the fan and filter element is reduced, prolonging the service life of the whole machine and reducing the maintenance cost. BRIEF DESCRIPTION OF DRAWINGS

[0031] Figure 1 The filtering performance optimization system based on the CFAU multi-module provided by the present application is provided with a structure general block diagram.

[0032] Figure 2 The filtering performance optimization system based on the CFAU multi-module provided by the present application is provided with a structure general block diagram.

[0033] Figure 3 The filtering performance optimization system based on the CFAU multi-module provided by the present application is provided with a structure general block diagram. DETAILED DESCRIPTION

[0034] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0035] In order to enable those skilled in the art to better understand the present application, the present application will be further described in detail below with reference to the drawings and specific embodiments.

[0036] Please refer to Figures 1 to 3, the filtering performance optimization system based on the CFAU multi-module, the system comprising: a C-Unit acquisition module for acquiring dust concentration data at the front end of the filter element and the rear end of the filter element, synchronously acquiring current data of the power supply line, and stamping the dust concentration data and the current data with a unified timestamp; performing adaptive interval processing on the dust concentration data collected in a preset time window at the front end of the filter element, to dynamically adjust the number of particle size intervals at the front end of the filter element and the rear end of the filter element and the particle size count vector according to the dust concentration fluctuation degree.

[0037] A F-Unit feedback evaluation module: for acquiring the multi-point differential pressure at the rear end of the filter element within a preset evaluation period and forming a differential pressure grid data, and generating a filter element partition load matrix, an instantaneous filtration efficiency and a unit air volume energy consumption based on the dust concentration data, the current data and the differential pressure grid data.

[0038] An A-Unit adaptive decision module: for generating a candidate control scheme according to the preset lower limit of filtration efficiency and the upper limit of unit air volume energy consumption of the system; adopting a double-threshold negotiation mechanism to output an execution instruction only when the candidate scheme meets the filtration efficiency and energy consumption constraints at the same time, and otherwise outputting a downshift instruction.

[0039] A U-Unit execution update module: for receiving the execution instruction to adjust the fan speed, switch the air duct or adjust the electrostatic field strength, and feeding back the energy consumption margin to the A-Unit adaptive decision module after execution.

[0040] The present application constructs four-level processes of acquisition (C-Unit), feedback (F-Unit), decision (A-Unit) and execution (U-Unit), takes the high-frequency multi-point sampling of C-Unit as the entrance, stamps the dust concentration before and after the filter element and the fan current with a timestamp and performs adaptive interval processing to reduce the data volume before inputting F-Unit; F-Unit quantifies the filter element partition load, the instantaneous filtration efficiency and the unit air volume energy consumption at the same time in each evaluation period and sends them to A-Unit; A-Unit quickly selects the best candidate control scheme that meets the requirements of “efficiency up to standard and energy consumption controlled” based on the efficiency-energy consumption double-threshold negotiation mechanism, and then issues the corresponding instruction to U-Unit; U-Unit executes the real-time adjustment of the fan speed, the air duct passage or the electrostatic field strength, and then feeds back the latest energy consumption margin to A-Unit, so as to promote the real-time iteration of the decision parameters in the next cycle. The present application not only realizes the adaptive compression and fine monitoring of multi-particle size particles, but also takes into account the filter element load balancing, energy consumption peak suppression and filter element life extension, and significantly improves the intelligent level and operation economy of the air filtration device.

[0041] In the embodiment of the present application, reference is made to Figure 1 and Figure 2 , as shown in Figure 1A filter performance optimization system structure total block diagram based on a CFAU multi-module is provided for the present application. Figure 2 A filter performance optimization system core workflow schematic diagram based on a CFAU multi-module is provided for the present application. In the present example, the filter performance optimization system based on a CFAU multi-module comprises the following modules.

[0042] A C-Unit acquisition module is used to acquire dust concentration data at the front end of the filter element and the rear end of the filter element, synchronously acquire current data of the power supply line, and stamp a unified timestamp on the dust concentration data and the current data; adaptive interval processing is performed on the dust concentration data acquired within a preset time window at the front end of the filter element, so as to dynamically adjust the number of particle size intervals at the front end of the filter element and the rear end of the filter element and the particle size count vector according to the dust concentration fluctuation degree.

[0043] In the embodiment of the present application, a group of laser particle counters is arranged at the front end of the filter element and the rear end of the filter element, and the count values of a multi-dimensional particle size channel (such as a 6-dimensional particle size channel, including ) are synchronously sampled at a preset frequency (such as 1 Hz), and the current of the fan power supply line is read in real time by a current detection board; all original data are first written into an on-chip ring buffer, and the MCU uniformly writes a timestamp at the moment of writing. The MCU extracts the total concentration sequence of the front end in a preset time window (such as 30 seconds) in a sliding window manner, and calculates the variance. When the variance is lower than a preset threshold, three segment interval count vectors such as are added respectively to form a three-dimensional structure; otherwise, the 6-dimensional structure remains unchanged. The determination result is applied to the vector reorganization at the rear end of the filter element, and a data structure with the same length is formed.

[0044] An F-Unit feedback evaluation module is used to acquire the multi-point differential pressure at the rear end of the filter element within a preset evaluation period and form a differential pressure grid data, and generate a filter partition load matrix, an instantaneous filtration efficiency and a unit air volume energy consumption based on the dust concentration data, the current data and the differential pressure grid data.

[0045] In the embodiment of the present application, a MEMS differential pressure probe (such as 4*4 position arrangement) is arranged at the rear end of the filter element, and the differential pressure grid data is generated by sampling once every preset evaluation period (such as 120 seconds); after the F-Unit receives continuous current values (such as 120) and the reorganized particle size count vectors, the overall grid differential pressure is first obtained by the row-column average method for the 4*4 differential pressure grid data, and compared with the filter reference differential pressure The grid load ratio is calculated, and the 4*4 filter partition load matrix is output; then the instantaneous filtration efficiency is obtained by using the front-end and rear-end particle size vectors, and the overall grid differential pressure is compared with the real-time current estimated unit air volume energy consumption ; in the formula, is the power supply voltage of the filter device, is the current collected by the C-Unit in real time, is the power factor of the filter device (a factory calibration constant), is the rated pressure difference, a proportional coefficient , is the rated air volume.

[0046] The A-Unit adaptive decision module is used to generate a candidate control scheme according to the system preset lower limit of filtration efficiency and upper limit of unit air volume energy consumption; a double threshold negotiation mechanism is adopted to output an execution instruction only when the candidate scheme meets the filtration efficiency and energy consumption constraints at the same time, otherwise a downshift instruction is output.

[0047] In the embodiment of the application, based on the current filter element partition load matrix, the instantaneous filtration efficiency gap and the energy consumption margin , a built-in efficiency-energy consumption approximation model is called to generate a number of candidate control schemes. Each scheme will get a predicted efficiency improvement and energy consumption increase ; the double threshold comparison unit confirms that the scheme is feasible only when and , and selects the optimal scheme according to the priority of the smallest efficiency gap and the smallest energy consumption increase, and outputs the corresponding execution instruction; if there is no feasible scheme, a downshift instruction is output and a negotiation timing window is started.

[0048] The U-Unit execution update module is used to receive the execution instruction to adjust the fan speed, switch the air duct or adjust the electrostatic field strength, and feedback the energy consumption margin to the A-Unit adaptive decision module after execution.

[0049] In the embodiment of the application, by receiving the execution instruction, the PWM control interface of the fan drive, the air valve stepper motor and the high-voltage electrostatic generator is called, the target RPM, valve angle and field strength are set according to the instruction, and the steady-state maintenance operation is entered; then the MCU re-samples the current and calculates the new unit air volume energy consumption, then calculates the new energy consumption margin and feeds back to the A-Unit, closes the next control cycle.

[0050] Preferably, the C-Unit acquisition module comprises a dust concentration acquisition unit for acquiring dust concentration data at the front end of the filter element and the rear end of the filter element at a preset sampling frequency.

[0051] The current detection unit is used to synchronously acquire the current data of the power supply circuit.

[0052] The time synchronization unit is used to timestamp dust concentration data and current data.

[0053] The adaptive interval unit is used to statistically analyze the dust concentration data within a preset time window at the front end of the filter element, dynamically adjust the number of particle size intervals at the front and rear ends of the filter element according to the degree of dust concentration fluctuation, and reorganize the particle size counting vector according to the adjusted number of particle size intervals.

[0054] In this embodiment of the invention, by installing a laser particle counter at both the front and rear ends of the filter element and uniformly setting a preset sampling frequency (1Hz) by the MCU, the dust concentration acquisition unit acquires multi-dimensional particle size channels (such as 6-dimensional particle size channels, including...) in real time. The count value is calculated; the current detection unit uses a Hall effect current sensor to synchronously measure the current of the fan power supply bus and sends the instantaneous current value and particle count together to the on-chip DMA buffer; the time synchronization unit timestamps the two raw data streams using a millisecond-level system clock. Then it is written to the circular buffer, forming a ring like this. The particle size count vector of the structure is then used, and the variance of the front-end total concentration sequence is calculated by adaptive interval cells within a preset time window (30 seconds). ,like Below the preset variance threshold Then count the particle size. Adjacent particle size ranges merge into a shape like The 3D particle size counting vector is used, thus halving the number of particle size intervals; otherwise, the 6D vector structure remains unchanged. This determination result is broadcast to the backend counting vector in real time to ensure that the dimensions of the two measurement points are consistent; finally, the reorganized particle size counting vector and current value are sent to the F-Unit to provide low-latency, structurally consistent input data for subsequent load assessment and efficiency-energy consumption calculation.

[0055] Preferred options, please refer to Figure 3 , Figure 3 This is a schematic diagram of the dust concentration data structure processing flow of the CFAU multi-module filtration performance optimization system of the present invention; the adaptive interval unit includes: a variance calculation subunit, used to calculate the variance of the dust concentration data within a preset time window at the front end of the filter element.

[0056] The threshold comparison subunit is used to compare the variance with a preset variance threshold. If the variance is less than the preset variance threshold, the number of particle size intervals is halved. If the variance is greater than or equal to the threshold, the number of particle size intervals remains unchanged.

[0057] The interval merging subunit is used to merge adjacent intervals in order and add the count values ​​of two adjacent intervals when the number of particle size intervals is halved, so as to obtain the reorganized particle size count vector at the front end and rear end of the filter element.

[0058] In this embodiment of the invention, the variance of the total concentration sequence at the front end of the filter cartridge within a preset time window (30 seconds) is calculated by a variance calculation subunit. Then the variance Input threshold comparison sub-unit and preset variance threshold comparison, when the variance Less than the preset variance threshold Time output marker At this point, the particle size counting vector will be... The particle size count vector is obtained by merging adjacent intervals in sequence and adding the count values ​​of two adjacent particle size intervals. Otherwise output In this case, the dimension remains unchanged. For example, when the sampling frequency is 1 Hz, there are 30 sampling points within 30 seconds, and each sampling point is a 6-dimensional particle size counting vector. When the variance of 30 sampling points is less than the preset variance threshold, they will be merged into one. The 3D particle size counting vector. When the dimension of the front end of the filter element is halved, the dimension of the back end of the filter element is also halved.

[0059] Preferably, the F-Unit feedback evaluation module includes: a differential pressure grid acquisition unit, used to collect multi-point differential pressure at the back end of the filter element according to a preset sampling period and form differential pressure grid data.

[0060] The load matrix calculation unit is used to compare the differential pressure grid data with the filter element reference differential pressure to generate the filter element partition load matrix.

[0061] The filtration efficiency calculation unit is used to calculate the real-time filtration efficiency based on the particle size count vectors at the front and rear ends of the reformed filter element.

[0062] The energy consumption calculation unit is used to calculate the energy consumption per unit air volume based on the average pressure difference between the power supply line current data and the pressure difference grid data.

[0063] In this embodiment of the invention, by uniformly arranging 4×4 MEMS differential pressure probes at the rear end of the filter element and triggering sampling by polling with the internal clock of the F-Unit at a preset sampling period (120 seconds) as an evaluation period, the differential pressure grid acquisition unit processes the raw readings of 16 measuring points. After bus aggregation, a 4×4 differential pressure grid is formed; subsequently, the load matrix calculation unit calculates the differential pressure based on the calibrated filter element reference pressure. Perform the following calculations for each grid cell: ;in, The differential pressure grid data values ​​are used to obtain the 4×4 filter element zone load matrix. The overall average pressure drop is calculated simultaneously using the following formula: .

[0064] Meanwhile, the F-Unit reads the front-end particle size count vector of the C-Unit just after the C-Unit is reformed and the rear-end particle size count vector The filtering efficiency calculation unit calculates the real-time filtering efficiency according to the following formula: .

[0065] Finally, the energy consumption calculation unit takes the real-time current and the filtering device supply voltage to calculate the fan power: ; wherein, is the filtering device power factor (a factory calibration constant), and then combines the proportional coefficient ( is the rated air volume) to calculate the instantaneous air volume , so as to obtain the unit air volume energy consumption .

[0066] Preferably, the A-Unit adaptive decision module comprises: a control scheme generation unit, configured to combine at least one of the fan gear, the air duct channel and the electrostatic field intensity according to a preset control rule to form a candidate control scheme, and estimate the efficiency improvement amount and the energy consumption increase amount corresponding to the candidate control scheme.

[0067] A double-threshold comparison unit is configured to compare the efficiency improvement amount with the lower limit of the system filtering efficiency, and compare the energy consumption increase amount with the upper limit of the current unit air volume energy consumption.

[0068] An instruction output unit is configured to output a candidate control scheme execution instruction when both thresholds are satisfied, and output a gear down instruction when any threshold is not satisfied.

[0069] In the embodiment of the application, the control scheme generation unit performs step search on the three-dimensional parameter space according to the preset control rule, and calls the calibrated efficiency-energy consumption approximation model to quickly output the efficiency improvement amount and the energy consumption increase amount for each group of parameters, thereby forming a candidate scheme list; the double-threshold comparison unit reads the candidate scheme list item by item, compares the lower limit of the system filtering efficiency and the upper limit of the unit air volume energy consumption , and only when and are satisfied, the scheme is retained; if the retained list is not empty, the first record is selected according to the priority of the minimum efficiency gap and the minimum energy consumption increase; if all candidate schemes do not satisfy the double thresholds, a gear down instruction is output and a negotiation timing window is started, and the rule deduction is triggered again after the window ends.

[0070] Preferably, the preset control rule comprises: when the number of high load blocks in the filter core partition load matrix exceeds a first preset threshold, a candidate control scheme of switching the air duct channel is preferentially generated.

[0071] When the instant filtration efficiency gap is less than a preset efficiency compensation threshold and the energy consumption margin is greater than a second preset threshold, a control scheme of increasing the electrostatic field intensity is preferentially generated.

[0072] When none of the above conditions is met, a control scheme of increasing the fan speed is generated.

[0073] In the embodiment of the application, by performing threshold scanning on the filter core partition load matrix , the number of high load grids satisfying is counted, if the number of high load grids is greater than a first preset threshold, it is determined that the local clogging of the filter core has been significant, the current channel is switched to the standby channel immediately by calling the air duct switching table, and the current channel is pushed to the first position of the scheme queue, otherwise, the instant filtration efficiency gap and the energy consumption margin are calculated. When the instant filtration efficiency gap is less than a preset efficiency compensation threshold and the energy consumption margin is greater than a second preset threshold, the electrostatic field intensity is increased on the premise that the fan speed is unchanged, and the fine particle capture is mainly subdivided. If the above two conditions are not met, the fan speed is enumerated from bottom to top. The three types of schemes are written into the candidate pool in the above order, and then enter the efficiency-energy consumption estimation and double-threshold cutting process to ensure that the final output can solve the clogging or insufficient efficiency without exceeding the upper limit of energy consumption.

[0074] Preferably, the control scheme generation unit calls an efficiency-energy consumption approximate model based on experimental calibration in a table lookup manner to estimate the efficiency improvement amount and the energy consumption increase amount of each candidate scheme.

[0075] In the embodiment of the application, a three-dimensional table model is pre-set, which is derived from experiments on different fan speeds, air duct valve positions and electrostatic field intensity combinations in a constant temperature clean room during factory calibration, and 120 groups of measured filtration efficiency-energy consumption pairs are obtained. For any candidate scheme generated during operation, the control scheme generation unit first maps the three control quantities to the table index interval respectively, and then uses a three-linear interpolation function to retrieve the efficiency improvement amount and the energy consumption increase amount at one time; if the candidate point falls on the calibration grid node, the measured value is directly returned without interpolation.

[0076] Preferably, the double-threshold comparison unit compares by the following relationship: and ; wherein, lower limit of system filtration efficiency, instant filtration efficiency, efficiency improvement of candidate control scheme current unit air volume energy consumption, energy consumption increase of candidate control scheme, upper limit of current unit air volume energy consumption.

[0077] It is noted that, in this document, phrases such as first and second should be construed in a non-limiting manner, merely to distinguish one entity or action from another, without necessarily implying that the entities or actions are in any way mutually exclusive or in any way in a sequence. Moreover, the terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus. An element proceeded by "comprises... a" does not, without more constraints, exclude the existence of additional identical elements in the process, method, article, or apparatus that comprises the element.

[0078] While embodiments of the present application have been shown and described with reference to a few embodiments, it will be understood by those skilled in the art that various changes in form and details can be made therein without departing from the spirit and scope of the application as defined by the appended claims and their equivalents.

Claims

1. A filter performance optimization system based on CFAU multi-module, characterized in that, The system comprises: a C-Unit acquisition module, configured to acquire dust concentration data at the front end and the rear end of the filter element, synchronously acquire current data of the power supply circuit, and stamp the dust concentration data and the current data with a unified timestamp; perform adaptive interval processing on the dust concentration data acquired within a preset time window at the front end of the filter element, so as to dynamically adjust the number of particle size intervals at the front end and the rear end of the filter element and reform the particle size count vector according to the fluctuation degree of the dust concentration; an F-Unit feedback evaluation module: configured to acquire multi-point pressure differences at the rear end of the filter element within a preset evaluation period and form pressure difference grid data, and generate a filter element partition load matrix, instantaneous filtration efficiency and unit air volume energy consumption based on the dust concentration data, the current data and the pressure difference grid data; an A-Unit adaptive decision module: configured to generate a candidate control scheme according to a preset lower limit of filtration efficiency and an upper limit of unit air volume energy consumption; adopt a double-threshold negotiation mechanism to output an execution instruction only when the candidate scheme meets the filtration efficiency and energy consumption constraints at the same time, otherwise output a downshift instruction; a U-Unit execution and update module: configured to receive the execution instruction to adjust the fan speed, switch the air duct or adjust the electrostatic field strength, and feed back the energy consumption margin to the A-Unit adaptive decision module after execution.

2. The CFAU multi-modal based filtration performance optimization system of claim 1, wherein, The C-Unit acquisition module comprises: a dust concentration acquisition unit, configured to acquire dust concentration data at the front end and the rear end of the filter element at a preset sampling frequency; a current detection unit, configured to synchronously acquire current data of the power supply circuit; a time synchronization unit, configured to timestamp the dust concentration data and the current data; an adaptive interval unit, configured to statistically analyze the dust concentration data within a preset time window at the front end of the filter element, dynamically adjust the number of particle size intervals at the front end and the rear end of the filter element according to the fluctuation degree of the dust concentration, and reform the particle size count vector according to the adjusted number of particle size intervals.

3. The CFAU multi-modal module based filtration performance optimization system as claimed in claim 2, wherein, The adaptive interval unit comprises: a variance calculation subunit, configured to calculate the variance of the dust concentration data within a preset time window at the front end of the filter element; a threshold comparison subunit, configured to compare the variance with a preset variance threshold, if the variance is less than the preset variance threshold, determine that the number of particle size intervals is halved, if the variance is greater than or equal to the threshold, keep the number of particle size intervals unchanged; an interval merging subunit, configured to, when it is determined that the number of particle size intervals is halved, pair and merge adjacent intervals in order and add the count values of the adjacent two particle size intervals to obtain the reformed particle size count vector at the front end and the rear end of the filter element.

4. The CFAU multi-modal module based filtration performance optimization system as claimed in claim 1, wherein, The F-Unit feedback evaluation module comprises: a pressure difference grid acquisition unit, configured to acquire multi-point pressure differences at the rear end of the filter element according to a preset sampling period and form pressure difference grid data; a load matrix calculation unit, configured to compare the pressure difference grid data with a filter element reference pressure difference to generate a filter element partition load matrix; a filtration efficiency calculation unit, configured to calculate the instantaneous filtration efficiency according to the reformed particle size count vector at the front end and the rear end of the filter element; an energy consumption calculation unit, configured to calculate the unit air volume energy consumption according to the current data of the power supply circuit and the average pressure difference value of the pressure difference grid data.

5. The CFAU multi-modal module based filtration performance optimization system as claimed in claim 1, wherein, The A-Unit adaptive decision module comprises: a control scheme generation unit configured to combine at least one of a fan gear, an air duct channel and an electrostatic field intensity according to a preset control rule to form a candidate control scheme, and estimate an efficiency improvement amount and an energy consumption increase amount corresponding to the candidate control scheme; a double-threshold comparison unit configured to compare the efficiency improvement amount with a system filtration efficiency lower limit, and compare the energy consumption increase amount with a current unit air volume energy consumption upper limit; and an instruction output unit configured to output a candidate control scheme execution instruction when the double-threshold comparison unit satisfies both conditions, and output a gear down instruction when either threshold does not satisfy the condition.

6. The CFAU multi-module based filtration performance optimization system as claimed in claim 5, wherein, The preset control rule comprises: when the number of high load blocks in the filter core partition load matrix exceeds a first preset threshold, a candidate control scheme of switching the air duct channel is preferentially generated; when the instant filtration efficiency gap is less than a preset efficiency compensation threshold and the energy consumption margin is greater than a second preset threshold, a control scheme of improving the electrostatic field intensity is preferentially generated; and when neither condition is satisfied, a control scheme of improving the fan gear is generated.

7. The CFAU multi-module based filtration performance optimization system as claimed in claim 5, wherein: The control scheme generation unit calls an efficiency-energy consumption approximate model established based on experimental calibration in a table lookup manner to estimate the efficiency improvement amount and the energy consumption increase amount of each candidate scheme.

8. The CFAU multi-modal based filtration performance optimization system of claim 5, wherein: The double-threshold comparison unit starts a negotiation timing window after outputting the gear down instruction, and re-triggers the control scheme generation unit to generate a new control scheme candidate and repeat the comparison process of the double-threshold comparison unit after the negotiation timing window expires.

9. The CFAU multi-module based filtration performance optimization system as claimed in claim 5, wherein, The double-threshold comparison unit compares by the following relationship: ; wherein, is the lower limit of the system filtration efficiency, is the instantaneous filtration efficiency, is the efficiency improvement amount of the candidate control scheme is the current unit wind volume energy consumption, is the energy consumption increase amount of the candidate control scheme, is the upper limit of the current unit wind volume energy consumption.

Images