Clean room air compression system energy consumption system construction method and device

Abstract

This application proposes a method and apparatus for constructing an energy consumption system for a cleanroom compressed air system. The method involves acquiring data such as pressure fluctuations, positive pressure deviation, sound energy loss of the dust collector fan, and the switching frequency of the air conditioning unit's heating and cooling coils. Dynamic energy consumption of the pipeline network is calculated using an LSTM time-series prediction model to determine the airflow balance adjustment coefficient. Based on the switching frequency of the heating and cooling coils, a critical temperature value for coil function switching is generated. Noise reduction gradient parameters are determined based on the sound energy loss value. The fresh air volume is controlled using the adjustment coefficient to generate real-time positive pressure compensation data. A heating and cooling coil switching instruction set is generated based on the critical temperature value. Noise reduction parameters are deployed, and acoustic parameters for the soundproof wall panels are constructed. These data and parameters are comprehensively monitored to generate energy efficiency improvement indicators. This method comprehensively optimizes multiple parameters of the compressed air system, effectively solving problems such as high energy consumption, unstable positive pressure, and high noise levels, significantly improving the energy efficiency and overall performance of the cleanroom system.

Description

Technical Field

[0001] This application belongs to the field of air compressor energy consumption, and in particular relates to a method and device for constructing an energy consumption system for a cleanroom air compressor system. Background Technology

[0002] In cleanroom systems, compressed air systems account for a significant proportion of energy consumption. Traditional compressed air systems lack intelligent energy management, leading to substantial energy waste. Meanwhile, positive pressure control in cleanrooms is crucial for maintaining the cleanliness of the indoor environment, but traditional positive pressure control methods often rely on manual experience, making precise control difficult. Furthermore, the noise from dust collector fans is a challenge in cleanroom systems; high noise levels not only affect staff comfort but can also negatively impact equipment stability. Additionally, the switching between hot and cold coils in the outdoor air conditioning unit is a crucial factor affecting energy consumption and stability; traditional switching methods lack intelligence, resulting in high energy consumption and insufficient stability. Summary of the Invention

[0003] The purpose of this application is to overcome the deficiencies in the prior art and provide a method and apparatus for constructing an energy consumption system for a cleanroom compressed air system.

[0004] This application provides a method for constructing an energy consumption system for a cleanroom compressed air system, including:

[0005] Acquire pressure fluctuation data of the cleanroom compressed air system, cleanroom positive pressure deviation sequence, dust removal fan sound energy loss value, and switching frequency of the cooling and heating coils of the outdoor air conditioning unit;

[0006] Based on the pressure fluctuation data, the dynamic energy consumption prediction data of the pipeline network is calculated using the LSTM time series prediction model.

[0007] Based on the network dynamic energy consumption prediction data and the cleanroom positive pressure deviation sequence, the air volume balance adjustment coefficient is determined.

[0008] Based on the switching frequency of the hot and cold coils in the outdoor air conditioning unit, a critical temperature value for coil function switching is generated through a proportional two-way valve.

[0009] Based on the sound energy loss value of the dust removal fan, the noise reduction gradient parameters of the double volute resonant cavity are determined;

[0010] The fresh air volume of MIAU is controlled by the air volume balance adjustment coefficient to generate real-time positive pressure compensation data;

[0011] Based on the critical temperature value for coil function conversion, a set of instructions for switching between hot and cold coils is generated.

[0012] Deploy the aforementioned noise reduction gradient parameters to construct the acoustic parameters of the sound insulation panel assembly;

[0013] A cleanroom control system is generated based on the real-time positive pressure compensation data, the hot and cold coil switching instruction set, and the acoustic parameters of the soundproof wall panel assembly.

[0014] Optionally, the LSTM time series prediction model includes: employing a sliding time window mechanism;

[0015] The steps of the sliding time window mechanism include: dividing the pressure fluctuation data into time windows according to a 24-hour cycle, constructing a pipeline energy consumption time series training set with a step size of 15 minutes, and outputting dynamic energy consumption prediction data for the next 2 hours.

[0016] Optionally, based on the network dynamic energy consumption prediction data and the cleanroom positive pressure deviation sequence, an air volume balance adjustment coefficient is determined, including:

[0017] The NSGA-II multi-objective processing algorithm is used to simultaneously optimize the three objective functions of air compressor unit energy consumption index, clean room positive pressure stability index, and equipment lifespan decay index to obtain the air volume balance adjustment coefficient.

[0018] Optionally, when the proportional two-way valve generates the critical temperature value for coil function switching, it includes:

[0019] When the outside air temperature is detected to be ≤5℃, the opening of the two-way valve is adjusted to 30-45% through the PID control algorithm so that the water flow rate of the cold coil is not less than 1.2m / s.

[0020] Optionally, based on the sound energy loss value of the dust collector fan, the noise reduction gradient parameters of the double volute resonant cavity are determined, including: setting a 30-100kg / m³ gradient density glass fiber layer on the inner side of the dust collector fan volute and integrating a Helmholtz resonator array, with the resonance frequency covering the 500-4000Hz noise frequency band.

[0021] This application also provides a device for constructing an energy consumption system for a cleanroom compressed air system, comprising:

[0022] The acquisition module acquires pressure fluctuation data of the cleanroom air compressor system, cleanroom positive pressure deviation sequence, dust removal fan sound energy loss value, and outdoor air conditioning unit cooling and heating coil switching frequency.

[0023] The pipeline module calculates the pipeline dynamic energy consumption prediction data using an LSTM time series prediction model based on the pressure fluctuation data.

[0024] The air volume module determines the air volume balance adjustment coefficient based on the network dynamic energy consumption prediction data and the cleanroom positive pressure deviation sequence;

[0025] The conversion module generates a critical temperature value for coil function conversion based on the switching frequency of the hot and cold coils of the outdoor air conditioning unit through a proportional two-way valve.

[0026] The noise reduction module determines the noise reduction gradient parameters of the double volute resonant cavity based on the sound energy loss value of the dust collector fan.

[0027] The compensation module controls the fresh air volume of the MIAU through the air volume balance adjustment coefficient and generates real-time positive pressure compensation data;

[0028] The instruction module generates a set of instructions for switching between hot and cold coils based on the critical temperature value for coil function conversion.

[0029] The acoustic module deploys the noise reduction gradient parameters to construct the acoustic parameters of the sound insulation panel assembly;

[0030] The indicator module generates a cleanroom control system based on the real-time positive pressure compensation data, the hot and cold coil switching instruction set, and the acoustic parameters of the soundproof wall panel assembly.

[0031] Optionally, the LSTM time series prediction model includes: employing a sliding time window mechanism;

[0032] The steps of the sliding time window mechanism include: dividing the pressure fluctuation data into time windows according to a 24-hour cycle, constructing a pipeline energy consumption time series training set with a step size of 15 minutes, and outputting dynamic energy consumption prediction data for the next 2 hours.

[0033] Optionally, the airflow module determines the airflow balance adjustment coefficient based on the network dynamic energy consumption prediction data and the cleanroom positive pressure deviation sequence, including:

[0034] The NSGA-II multi-objective processing algorithm is used to simultaneously optimize the three objective functions of air compressor unit energy consumption index, clean room positive pressure stability index, and equipment lifespan decay index to obtain the air volume balance adjustment coefficient.

[0035] Optionally, when the proportional two-way valve in the conversion module generates the critical temperature value for coil function conversion, it includes:

[0036] When the outside air temperature is detected to be ≤5℃, the opening of the two-way valve is adjusted to 30-45% through the PID control algorithm so that the water flow rate of the cold coil is not less than 1.2m / s.

[0037] Optionally, the acoustic module determines the noise reduction gradient parameters of the double-volute resonant cavity based on the sound energy loss value of the dust collector fan, including:

[0038] A 30-100kg / m³ gradient density glass fiber layer is set inside the volute of the dust collector fan, and a Helmholtz resonator array is integrated, with the resonant frequency covering the 500-4000Hz noise band.

[0039] The beneficial effects of this application are:

[0040] Invention point:

[0041] 1. Design an LSTM neural network with a composite loss function for predicting dynamic energy consumption in pipeline networks.

[0042] 2. Multi-objective optimization equations and dynamic adjustment coefficient matrix based on energy-pressure dual constraints.

[0043] 3. Design multi-layer composite sound insulation wall panels, combining micro-perforated panels with damping treatment technology.

[0044] 4. Multi-dimensional energy efficiency improvement index calculation model.

[0045] This application provides a method for constructing an energy consumption system for a cleanroom compressed air system, comprising: acquiring pressure fluctuation data, a positive pressure deviation sequence, sound energy loss value of a dust removal fan, and the switching frequency of the hot and cold coils of an outdoor air conditioning unit; calculating dynamic energy consumption prediction data of the pipeline network using an LSTM time-series prediction model based on the pressure fluctuation data; determining an airflow balance adjustment coefficient based on the network dynamic energy consumption prediction data and the cleanroom positive pressure deviation sequence; generating a critical temperature value for coil function switching using a proportional two-way valve based on the switching frequency of the hot and cold coils of the outdoor air conditioning unit; determining the silencing gradient parameters of the double volute resonant cavity based on the sound energy loss value of the dust removal fan; controlling the MIAU fresh air volume using the airflow balance adjustment coefficient to generate real-time positive pressure compensation data; generating a hot and cold coil switching instruction set based on the critical temperature value for coil function switching; deploying the silencing gradient parameters to construct the acoustic parameters of the soundproof wall panel assembly; and generating a cleanroom control system based on the real-time positive pressure compensation data, the hot and cold coil switching instruction set, and the acoustic parameters of the soundproof wall panel assembly. This application, by comprehensively considering and optimizing multiple parameters of the air compressor system, effectively solves the problems of high energy consumption, unstable positive pressure control, and high noise in cleanroom air compressor systems, and significantly improves the energy efficiency and overall performance of cleanroom systems. Attached Figure Description

[0046] Figure 1 This is a schematic diagram of the process for constructing energy consumption indicators for the cleanroom compressed air system in this application;

[0047] Figure 2 This is a schematic diagram of the device for constructing energy consumption indicators for the cleanroom compressed air system in this application. Detailed Implementation

[0048] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it will be understood that various forms of implementation of the present disclosure are possible and should not be limited to the embodiments set forth herein. Rather, the embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0049] This application provides a method for constructing an energy consumption system for a cleanroom compressed air system, including:

[0050] S101. Obtain pressure fluctuation data of the cleanroom air compressor system, cleanroom positive pressure deviation sequence, dust removal fan sound energy loss value and outdoor air conditioning unit cooling and heating coil switching frequency.

[0051] Pressure fluctuation data acquisition: Pressure sensors are deployed at key nodes of the air compressor system, including the compressor outlet, air tank inlet and outlet, front and rear ends of the dryer, and front end of the end-use equipment, to collect pressure fluctuation data. The pressure sensors are set to a sampling frequency of 10Hz and transmit data via RS485 bus.

[0052] Using a moving average filter, for example with a window width of 50 sampling points, high-frequency noise is eliminated, and the standard deviation is calculated as a fluctuation index, as shown in the following expression:

[0053]

[0054] in, As a volatility indicator, Let i be the number of sampling points, and i be the number of sampling points. This is the pressure value. This represents the average pressure.

[0055] Positive pressure deviation sequence acquisition:

[0056] In each area of ​​the cleanroom, including the production area, buffer zone, and changing area, differential pressure transmitters are installed to collect differential pressure data and establish a positive pressure deviation sequence, expressed as follows:

[0057]

[0058] Where k is the number of samplings and p is the pressure difference.

[0059] Sound energy loss calculation:

[0060] Install IEC 61672 sound level meters on the inlet and outlet pipes of the dust collector fan to measure the sound pressure level in the 63Hz-8kHz frequency band and calculate the sound power.

[0061] = ×A×t

[0062] Where A is the cross-sectional area of ​​the pipe, and t is the measurement time. This refers to the sound pressure level.

[0063] Monitoring of hot and cold coil switching frequency:

[0064] An electromagnetic flow meter and a PT100 temperature sensor were installed in the water circuit of the air conditioning unit to record the temperature difference ΔT (°C) between the inlet and outlet water.

[0065] Switching frequency calculation:

[0066]

[0067] S102. Based on the pressure fluctuation data, calculate the dynamic energy consumption prediction data of the pipeline network using the LSTM time series prediction model;

[0068] A three-layer LSTM neural network was constructed, with the input layer containing 32 neurons corresponding to time-series parameters such as pressure, flow rate, and temperature. A time-series sample set was established with a 60-minute time window and a 5-minute sliding step. The loss function adopted was a combination of MAE (Mean Absolute Error) and MAPE (Mean Absolute Percentage Error).

[0069] Training dataset construction:

[0070] 576 training samples were generated in 15-minute increments over a 24-hour period; Min-Max normalization was used to the [0,1] interval; the composite loss function was (MAE + 0.3×MAPE).

[0071]

[0072] in, This represents the actual energy consumption value. To predict energy consumption, N is the number of samples.

[0073] By introducing a 24-hour periodic sliding window (15-minute step), the daily periodic component, trend component, and residual component of pipeline pressure can be separated. The sliding window mechanism is equivalent to performing a locally weighted regression on the periodic component, enhancing the model's ability to capture diurnal pressure fluctuation patterns.

[0074] Meanwhile, pipeline energy consumption prediction considers both absolute energy consumption deviation (MAE) and relative fluctuation ratio (MAPE). Specifically, when the energy consumption base is large, MAE dominates the optimization and may ignore small abnormal fluctuations, while MAPE can lead to gradient explosion during low energy consumption periods (such as nighttime). Therefore, a linear combination (MAE + 0.3MAPE) is used to ensure that the model maintains stable convergence under both high and low load conditions.

[0075] Model training and validation:

[0076] Supervised learning was performed using historical runtime data with the Adam optimizer set to an initial learning rate of 0.001 and a batch size of 128. Model performance was evaluated through cross-validation, requiring a test set RMSE ≤ 3.5%.

[0077] By balancing absolute and relative errors using a composite loss function (MAE + 0.3 × MAPE), the accuracy of the test set is effectively improved compared to the traditional single loss function model.

[0078] The sliding window mechanism enhances the ability to capture local features of time series data, effectively reducing the deviation between predicted values ​​and actual energy consumption.

[0079] S103. Based on the network dynamic energy consumption prediction data and the cleanroom positive pressure deviation sequence, determine the air volume balance adjustment coefficient;

[0080] First, we model a multi-objective optimization problem:

[0081] If the decision variable is set to the air supply volume Qsupply, then the objective function is:

[0082]

[0083]

[0084]

[0085] Constraints: 2000≤Qsupply≤5000 (upper and lower limits of air volume).

[0086] in, Let the objective function be the daily energy consumption of the air compressor unit. Let the objective function be the mean value of the positive pressure deviation. Let the objective function be the equipment lifespan degradation rate. The predicted energy consumption value for hour t. Let be the positive pressure deviation of the i-th sampling.

[0087] NSGA-II algorithm parameter settings:

[0088] Population size: 100; crossover probability: 0.8; mutation probability: 0.1; maximum number of iterations: 200.

[0089] Dynamic weight allocation strategy:

[0090] Normal operating conditions: weights ω1=0.5, ω2=0.3, ω3=0.2.

[0091] When the positive pressure deviation is >5Pa: it will automatically adjust to ω1=0.3, ω2=0.5, ω3=0.2.

[0092] Pareto solution set filtering rules:

[0093] Retain non-dominated solutions that simultaneously satisfy the following conditions:

[0094]

[0095]

[0096]

[0097] in, This is the rated energy consumption.

[0098] Adjustment coefficient matrix:

[0099] Calculation of zoning adjustment coefficient: ,(i=1,2,...,n).

[0100] Matrix format: T, (n: number of air supply branches).

[0101] This application employs the NSGA-II algorithm to simultaneously optimize three objectives: energy consumption (f1, f2, f3), generating an air volume balance adjustment coefficient matrix K. The NSGA-II algorithm's Pareto solution set selection rules (f1, f2) ensure that both energy consumption and stability targets are met, while the optimization of the equipment lifespan degradation rate (f3) reduces the number of air compressor start-ups and shutdowns, increasing the expected lifespan.

[0102] The nonlinear coupling between f1 and f2 makes it difficult for traditional single-objective optimization to balance their contradictions. This application preserves the non-dominated solution set satisfying the constraints f1, f2, and f3 through non-dominated sorting, ensuring that the solution set simultaneously approximates the optimal values ​​of all three objectives, rather than compromising a single objective. Independent calculations are performed for different air supply branches. This solves the problem of local positive pressure imbalance caused by traditional global unified regulation.

[0103] S104. Based on the switching frequency of the hot and cold coils of the outdoor air conditioning unit, a critical temperature value for coil function conversion is generated through a proportional two-way valve.

[0104] The critical temperature calculation model includes: based on the heat balance equation, basic PID control, fuzzy rule base and low temperature protection mechanism.

[0105] Based on the thermal balance equation:

[0106]

[0107] =20℃ (set temperature)

[0108] =3℃ (safety margin)

[0109] Wall heat loss (measured in real time using an infrared thermal imager)

[0110] UA = 1.2 kW / ℃ (coil heat transfer coefficient × area)

[0111] PID-Fuzzy composite control algorithm

[0112] Basic PID control:

[0113] u(k) = 2.5e(k) + 0.8 +1.2 (Ts=5s)

[0114] Fuzzy rule base:

[0115] >3 +15 1-3 +5 -1~1 0 <-3 -15

[0116] Low temperature protection mechanism:

[0117] when At ≤5℃: Force the heating coil mode to open; limit the two-way valve opening to 30-45% to ensure a water flow velocity ≥1.2m / s; water flow velocity calculation formula:

[0118] v=

[0119] in, Volumetric flow rate, This represents the cross-sectional area of ​​the pipe.

[0120] Step S105: Determine the noise reduction gradient parameters of the double volute resonant cavity based on the sound energy loss value of the dust removal fan;

[0121] The design of Helmholtz resonators includes: individual resonator parameters, array arrangement scheme, and gradient density anechoic layer structure.

[0122] Individual resonator parameters:

[0123] =

[0124] In this context, d is the diameter of the neck opening, L is the neck length, and V is the cavity volume.

[0125] Array layout scheme:

[0126] 500-800 12 The inlet of the volute is evenly distributed around the circumference 800-2000 8 Spiral arrangement behind the impeller 2000-4000 6 Export diffuser section matrix arrangement

[0127] Gradient density sound-absorbing layer structure:

[0128] Material parameters:

[0129] Inner layer Fiberglass 80 30 Middle layer Ceramic cotton 50 20 outer layer Polyester fiber 30 10

[0130] Acoustic performance verification includes: insertion loss calculation, expressed as follows:

[0131] IL=10 ( ≥12dB, (500−4000Hz)

[0132] S106. Control the fresh air volume of MIAU through the air volume balance adjustment coefficient to generate real-time positive pressure compensation data;

[0133] Feedforward-feedback composite control model:

[0134] Compensation amount calculation:

[0135] ΔQ=0.35× +0.65×∫( - )dt

[0136] Dynamic limiting strategy:

[0137] >5 ±20% <10s 2-5 ±10% <20s <2 ±5% <30s

[0138] MIAU fresh air volume control execution:

[0139] PID parameter tuning:

[0140] =1.2, =0.5, =0.8 \quad.

[0141] Among them, the dead zone is ±2Pa and the overshoot is ≤5%.

[0142] In this application, feedforward control quickly offsets sudden pressure disturbances, feedback integral term eliminates steady-state error, dynamic limiting strategy prevents overshoot, and pressure recovery time is shortened.

[0143] S107. Generate a set of instructions for switching between hot and cold coils based on the critical temperature value for coil function conversion.

[0144] Based on the critical temperature value generated in step S104 Build dynamic switching logic:

[0145]

[0146] Set a lag bandwidth of 1℃ to avoid frequent switching (switching interval ≥ 10 minutes).

[0147] Control logic flow:

[0148] Read the outside air temperature and the critical temperature, and calculate the temperature difference deviation e= - ;

[0149] The valve position adjustment rate is determined by a fuzzy rule table (see step S104).

[0150] Generate instructions that include target opening degree, execution duration, and flow rate constraints.

[0151] Handling Abnormal Operating Conditions:

[0152] Water temperature ≤3℃ Forced switch to heat coil, start electric heat tracing system Water flow velocity <1.0m / s Shutting down the corresponding drive triggered an alarm. Switching more than 6 times within 1 hour Lock current mode for 2 hours

[0153] Instruction set deployment and execution:

[0154] Instruction storage format: Redis real-time database, sorted by timestamp; Execution cycle: Instruction queue is scanned every 30 seconds; Interface with BMS system: Control signals are sent to the air conditioning unit PLC via OPC UA protocol.

[0155] S108. Deploy the noise reduction gradient parameters to construct the acoustic parameters of the sound insulation wall panel assembly;

[0156] The soundproof wall panel adopts a three-layer structure: galvanized steel plate + centrifugal glass wool + micro-perforated aluminum plate, with a gradient density glass fiber layer of 80→50→30kg / m³ on the inner side, and a Helmholtz resonator array.

[0157] The Helmholtz resonator covers the 500-4000Hz frequency band with an insertion loss ≥12dB.

[0158] Soundproof wall panel structural parameters:

[0159] outer layer galvanized steel sheet 1.5mm + 2mm damping coating Constraint Layer Damping (CLD) Intermediate layer Centrifugal glass wool 50mm / 50kg / m³ Cover with non-woven fabric to prevent fiber shedding Inner layer Micro-perforated aluminum plate 0.8mm / Aperture 0.6mm Perforation rate 2%, hole spacing 15mm

[0160] Helmholtz resonator array deployment:

[0161] The resonant cavity design parameters are based on the calculation results of step S105:

[0162] 500-800Hz 12 8.2 Φ20×30 800-2000Hz 8 3.5 Φ15×25 2000-4000Hz 6 1.2 Φ10×20

[0163] Gradient density sound-absorbing layer:

[0164] Fiberglass layer density gradient: linearly decreasing from the inside to the outside at 80→50→30kg / m³; interlayer bonding: using high-temperature resistant silicone (shear strength ≥0.8MPa); joint treatment: board joint width ≤2mm, filled with polyurethane sealant (attenuation ≥3dB / joint).

[0165] Acoustic performance verification and calibration:

[0166] Test methods: Sound intensity scanning method (ISO 9614-2): Measured at grid points at a distance of 1m (0.5m interval); Reverberation chamber method (ISO 354): Measure the transmission loss (TL) of the sound insulation panel; Acceptance criteria: TLavg≥25dB ​​(500−4000Hz) ILmax≥12dB (target frequency band of the resonator).

[0167] In this application, the gradient density layer effectively absorbs mid-to-high frequency noise, the Helmholtz resonator reduces noise at low-frequency peaks, and the micro-perforated plate suppresses high-frequency airflow regeneration noise.

[0168] S109. Based on the real-time positive pressure compensation data, the hot and cold coil switching instruction set, and the acoustic parameters of the soundproof wall panel assembly, a cleanroom control system is generated.

[0169] Establish a multidimensional data matrix: X= Principal component analysis was used to extract eigenvectors, retaining components with a cumulative contribution rate ≥ 85%, where X is the original data matrix for energy efficiency assessment. This is a positive pressure deviation sequence. For real-time air supply volume, Sound pressure level, For the number of times the hot and cold coils are switched, This represents the average value of the air volume balance adjustment coefficient.

[0170] Construct a comprehensive evaluation function: η = 0.3 + 0.25· + 0.2· +0.15· + 0.1·MTBF, where For energy saving rate, This is the pressure stability index. For noise reduction, For cost savings, MTBF is the mean time between failures.

[0171] Energy efficiency =( - ) / ×100%; Pressure stability =1 - σ(ΔP) / μ(ΔP); Noise reduction amount = - Cost savings = + System reliability MTBF = runtime / number of failures; where, , These are the baseline energy consumption value and the actual energy consumption value, respectively, and σ(ΔP) and μ(ΔP) are the standard deviation and mean value of the positive pressure deviation, respectively. , These are the equivalent continuous A-weighted sound level before and after the modification, respectively. , These represent savings in initial investment and cumulative maintenance costs, respectively.

[0172] Traditional methods use only a single indicator (such as energy consumption), while this application adds other indicators. Furthermore, as described above, each indicator is determined through… and The covariance is adjusted in real time to effectively improve system energy efficiency and reduce costs.

[0173] In this application, principal component analysis is used to extract key features, avoiding the one-sidedness of a single indicator. Energy efficiency indicators are monitored in real time to provide early warning of equipment deterioration, and cost-saving indicators are quantified to reflect the initial investment and maintenance benefits.

[0174] This application also provides a device for constructing an energy consumption system for a cleanroom compressed air system, comprising:

[0175] Module 201 acquires pressure fluctuation data of the cleanroom air compressor system, cleanroom positive pressure deviation sequence, dust removal fan sound energy loss value, and outdoor air conditioning unit cooling and heating coil switching frequency.

[0176] Pipeline module 202 calculates dynamic energy consumption prediction data of the pipeline network based on the pressure fluctuation data using an LSTM time series prediction model;

[0177] The air volume module 203 determines the air volume balance adjustment coefficient based on the network dynamic energy consumption prediction data and the cleanroom positive pressure deviation sequence;

[0178] The conversion module 204 generates a critical temperature value for coil function conversion based on the switching frequency of the hot and cold coils of the outdoor air conditioning unit through a proportional two-way valve.

[0179] The noise reduction module 205 determines the noise reduction gradient parameters of the double volute resonant cavity based on the sound energy loss value of the dust removal fan.

[0180] The compensation module 206 controls the fresh air volume of MIAU through the air volume balance adjustment coefficient and generates real-time positive pressure compensation data;

[0181] Instruction module 207 generates a set of instructions for switching between hot and cold coils based on the critical temperature value for coil function conversion.

[0182] Acoustic module 208 deploys the noise reduction gradient parameters to construct the acoustic parameters of the sound insulation wall panel assembly;

[0183] The indicator module 209 generates a cleanroom control system based on the real-time positive pressure compensation data, the hot and cold coil switching instruction set, and the acoustic parameters of the soundproof wall panel assembly.

[0184] Furthermore, the LSTM time series prediction model includes: employing a sliding time window mechanism;

[0185] The steps of the sliding time window mechanism include: dividing the pressure fluctuation data into time windows according to a 24-hour cycle, constructing a pipeline energy consumption time series training set with a step size of 15 minutes, and outputting dynamic energy consumption prediction data for the next 2 hours.

[0186] Furthermore, the airflow module determines the airflow balance adjustment coefficient based on the network dynamic energy consumption prediction data and the cleanroom positive pressure deviation sequence, including:

[0187] The NSGA-II multi-objective processing algorithm is used to simultaneously optimize the three objective functions of air compressor unit energy consumption index, clean room positive pressure stability index, and equipment lifespan decay index to obtain the air volume balance adjustment coefficient.

[0188] Furthermore, when the proportional two-way valve in the conversion module generates the critical temperature value for coil function conversion, it includes:

[0189] When the outside air temperature is detected to be ≤5℃, the opening of the two-way valve is adjusted to 30-45% through the PID control algorithm so that the water flow rate of the cold coil is not less than 1.2m / s.

[0190] Furthermore, the acoustic module determines the noise reduction gradient parameters of the double-volute resonant cavity based on the sound energy loss value of the dust collector fan, including:

[0191] A 30-100kg / m³ gradient density glass fiber layer is set inside the volute of the dust collector fan, and a Helmholtz resonator array is integrated, with the resonant frequency covering the 500-4000Hz noise band.

[0192] The above description of the embodiments is provided to enable those skilled in the art to understand and apply the present invention. It will be apparent to those skilled in the art that various modifications can be made to the above embodiments, and the general principles described herein can be applied to other embodiments without inventive effort. Therefore, the present invention is not limited to the above embodiments, and any improvements and modifications made to the present invention by those skilled in the art based on the disclosure thereof should be within the scope of protection of the present invention.

Claims

1. A clean room air compression system energy consumption system construction method, characterized by, include: Acquire pressure fluctuation data of the cleanroom compressed air system, cleanroom positive pressure deviation sequence, dust removal fan sound energy loss value, and switching frequency of the cooling and heating coils of the outdoor air conditioning unit; Based on the pressure fluctuation data, the dynamic energy consumption prediction data of the pipeline network is calculated using the LSTM time series prediction model. Based on the predicted dynamic energy consumption data of the pipeline network and the positive pressure deviation sequence of the cleanroom, the air volume balance adjustment coefficient is determined. Based on the switching frequency of the hot and cold coils in the outdoor air conditioning unit, a critical temperature value for coil function switching is generated through a proportional two-way valve. Based on the sound energy loss value of the dust removal fan, the noise reduction gradient parameters of the double volute resonant cavity are determined; The fresh air volume of MIAU is controlled by the air volume balance adjustment coefficient to generate real-time positive pressure compensation data; Based on the critical temperature value for coil function conversion, a set of instructions for switching between hot and cold coils is generated. Deploy the aforementioned noise reduction gradient parameters to construct the acoustic parameters of the sound insulation panel assembly; Based on the real-time positive pressure compensation data, the hot and cold coil switching instruction set, and the acoustic parameters of the soundproof wall panel assembly, a cleanroom control system is generated; this includes, Building a multi-dimensional data matrix: X= ; using principal component analysis to extract the characteristic vector, retaining the cumulative contribution rate ≥85% of the components, wherein X is the original data matrix of energy efficiency evaluation, is a positive pressure deviation sequence, is a real-time air supply, is a sound pressure level, is the number of cold and hot coil switching, is the average value of air volume balance adjustment coefficient; Construct a comprehensive evaluation function: η=0.3 + 0.25 + 0.2 + 0.15 + 0.1MTBF; in For energy saving rate, This is the pressure stability index. For noise reduction, For cost savings, MTBF is the mean time between failures; Energy efficiency =( - ) / ×100%; Pressure stability =1 - σ(ΔP) / μ(ΔP); Noise reduction amount = - Cost savings = + System reliability MTBF = runtime / number of failures; where, , These are the baseline energy consumption value and the actual energy consumption value, respectively, and σ(ΔP) and μ(ΔP) are the standard deviation and mean value of the positive pressure deviation, respectively. , These are the equivalent continuous A-weighted sound level before and after the modification, respectively. , These represent savings in initial investment and cumulative maintenance costs, respectively.

2. The method for constructing an energy consumption system for a cleanroom compressed air system according to claim 1, characterized in that, The LSTM time series prediction model includes: employing a sliding time window mechanism; The steps of the sliding time window mechanism include: dividing the pressure fluctuation data into time windows according to a 24-hour cycle, constructing a pipeline energy consumption time series training set with a step size of 15 minutes, and outputting dynamic energy consumption prediction data for the next 2 hours.

3. The method for constructing an energy consumption system for a cleanroom compressed air system according to claim 1, characterized in that, Based on the predicted dynamic energy consumption data of the pipeline network and the positive pressure deviation sequence of the cleanroom, the air volume balance adjustment coefficient is determined, including: The NSGA-II multi-objective processing algorithm is used to simultaneously optimize the three objective functions of air compressor unit energy consumption index, clean room positive pressure stability index, and equipment lifespan decay index to obtain the air volume balance adjustment coefficient.

4. The method for constructing an energy consumption system for a cleanroom compressed air system according to claim 1, characterized in that, When the proportional two-way valve generates the critical temperature value for coil function conversion, it includes: When the outside air temperature is detected to be ≤5℃, the opening of the two-way valve is adjusted to 30-45% through the PID control algorithm so that the water flow rate of the cold coil is not less than 1.2m / s.

5. The method for constructing an energy consumption system for a cleanroom compressed air system according to claim 1, characterized in that, Based on the sound energy loss value of the dust collector fan, the noise reduction gradient parameters of the double volute resonant cavity are determined, including: setting a 30-100kg / m³ gradient density glass fiber layer on the inner side of the dust collector fan volute and integrating a Helmholtz resonator array, with the resonance frequency covering the 500-4000Hz noise frequency band.

6. A device for constructing an energy consumption system for a cleanroom compressed air system, characterized in that, include: The acquisition module acquires pressure fluctuation data of the cleanroom air compressor system, cleanroom positive pressure deviation sequence, dust removal fan sound energy loss value, and outdoor air conditioning unit cooling and heating coil switching frequency. The pipeline module calculates the pipeline dynamic energy consumption prediction data using an LSTM time series prediction model based on the pressure fluctuation data. The air volume module determines the air volume balance adjustment coefficient based on the dynamic energy consumption prediction data of the pipeline network and the positive pressure deviation sequence of the clean room; The conversion module generates a critical temperature value for coil function conversion based on the switching frequency of the hot and cold coils of the outdoor air conditioning unit through a proportional two-way valve. The noise reduction module determines the noise reduction gradient parameters of the double volute resonant cavity based on the sound energy loss value of the dust collector fan. The compensation module controls the fresh air volume of the MIAU through the air volume balance adjustment coefficient and generates real-time positive pressure compensation data; The instruction module generates a set of instructions for switching between hot and cold coils based on the critical temperature value for coil function conversion. The acoustic module deploys the noise reduction gradient parameters to construct the acoustic parameters of the sound insulation panel assembly; The indicator module generates a cleanroom control system based on the real-time positive pressure compensation data, the hot and cold coil switching instruction set, and the acoustic parameters of the soundproof wall panel assembly. This system includes... Establish a multidimensional data matrix: X= Principal component analysis was used to extract eigenvectors, retaining components with a cumulative contribution rate ≥ 85%, where X is the original data matrix for energy efficiency assessment. This is a positive pressure deviation sequence. For real-time air supply volume, Sound pressure level, For the number of times the hot and cold coils are switched, This represents the average value of the air volume balance adjustment coefficient. Construct a comprehensive evaluation function: η=0.3 + 0.25 + 0.2 + 0.15 + 0.1MTBF; in For energy saving rate, This is the pressure stability index. For noise reduction, For cost savings, MTBF is the mean time between failures; Energy efficiency =( - ) / ×100%; Pressure stability =1 - σ(ΔP) / μ(ΔP); Noise reduction amount = - Cost savings = + System reliability MTBF = runtime / number of failures; where, , These are the baseline energy consumption value and the actual energy consumption value, respectively, and σ(ΔP) and μ(ΔP) are the standard deviation and mean value of the positive pressure deviation, respectively. , These are the equivalent continuous A-weighted sound level before and after the modification, respectively. , These represent savings in initial investment and cumulative maintenance costs, respectively.

7. The cleanroom air compressor system energy consumption system construction device according to claim 6, characterized in that, The LSTM time series prediction model includes: employing a sliding time window mechanism; The steps of the sliding time window mechanism include: dividing the pressure fluctuation data into time windows according to a 24-hour cycle, constructing a pipeline energy consumption time series training set with a step size of 15 minutes, and outputting dynamic energy consumption prediction data for the next 2 hours.

8. The cleanroom compressed air system energy consumption system construction device according to claim 6, characterized in that, The airflow module determines the airflow balance adjustment coefficient based on the dynamic energy consumption prediction data of the pipeline network and the positive pressure deviation sequence of the cleanroom, including: The NSGA-II multi-objective processing algorithm is used to simultaneously optimize the three objective functions of air compressor unit energy consumption index, clean room positive pressure stability index, and equipment lifespan decay index to obtain the air volume balance adjustment coefficient.

9. The cleanroom compressed air system energy consumption system construction device according to claim 6, characterized in that, When the proportional two-way valve in the conversion module generates the critical temperature value for coil function conversion, it includes: When the outside air temperature is detected to be ≤5℃, the opening of the two-way valve is adjusted to 30-45% through the PID control algorithm so that the water flow rate of the cold coil is not less than 1.2m / s.

10. The cleanroom compressed air system energy consumption system construction device according to claim 6, characterized in that, The acoustic module determines the noise reduction gradient parameters of the double-volute resonant cavity based on the sound energy loss value of the dust collector fan, including: A 30-100kg / m³ gradient density glass fiber layer is set inside the volute of the dust collector fan, and a Helmholtz resonator array is integrated, with the resonant frequency covering the 500-4000Hz noise band.

Images