Modular building air tightness detection adaptive pressure closed loop intelligent control system

Abstract

The application discloses a modular building air tightness detection self-adaptive pressure closed loop intelligent control system and relates to the technical field of automatic pressure control; the system identifies the same pressure different flow return bifurcation caused by the critical uncovering and opening of the module joint gap through pressure regulation and return process, respectively establishes an uncovered and opened branch model and a return branch model after uncovering and opening, and adopts a multi-objective particle swarm algorithm to determine corresponding pressure control parameters; according to the branch state of the joint gap, the branch model and the control parameters are selected, air supply and exhaust execution instructions are generated, and the stable establishment of the target detection pressure is realized. The application can reduce the pressure overshoot, repeated fall and control quantity fluctuation caused by model mismatch, and improve the stability of the air tightness detection process.

Description

Technical Field

[0001] This invention relates to the field of automatic pressure control technology, and in particular to an adaptive pressure closed-loop intelligent control system for airtightness detection in modular buildings. Background Technology

[0002] With the development of prefabricated and modular buildings, building modules are assembled on-site through horizontal joints, vertical joints, and connection nodes. Their overall airtightness directly affects building energy conservation, indoor environmental stability, and reliability. Existing modular building airtightness testing typically uses fan pressurization or depressurization. By adjusting the supply and exhaust air volume, multiple preset test pressure differences are created between the interior and exterior of the building. After each test pressure difference stabilizes, the corresponding airflow is collected to obtain the pressure-flow relationship of the building envelope. To improve testing efficiency, existing testing equipment typically uses proportional-integral control, graded pressure setting, or fixed pressure change rate to perform closed-loop regulation of fan speed or electric damper opening.

[0003] However, when the differential pressure at the module joint increases step by step, it may exceed the critical opening pressure, causing the equivalent leakage state of the joint sealing interface to change. When the differential pressure subsequently decreases and returns to the previous pressure range, the joint may not recover immediately due to friction, delayed rebound of the sealing material, or constraints of the connection structure, resulting in a significant difference between the return air flow and the pressure regulating air flow under the same differential pressure. Existing control methods usually use a single pressure-flow model and fixed control parameters, which makes it difficult to identify the above-mentioned return states with the same pressure but different flow. It is easy to misjudge the pressure drop caused by the change in leakage state as insufficient air supply and exhaust capacity, and then continuously increase the control output, resulting in pressure overshoot, repeated drops, prolonged pressure stabilization time, and frequent changes in the control commands of the fan or air valve. Summary of the Invention

[0004] The purpose of this invention is to solve the problem in the prior art that the same pressure and different flow return occurs after the critical opening of the modular joint. The single pressure-flow model and fixed control parameters are difficult to adapt to the equivalent leakage state changes, which easily causes the detection pressure to overshoot, repeatedly drop, prolong the pressure stabilization time and the frequent fluctuation of the supply and exhaust air control commands. The proposed invention is an adaptive pressure closed-loop intelligent control system for modular building airtightness detection.

[0005] To address the problems existing in the prior art, the present invention adopts the following technical solution:

[0006] The modular building airtightness detection adaptive pressure closed-loop intelligent control system includes:

[0007] The system collects internal building pressure, external reference pressure, and effective supply and exhaust air volume to obtain the detection pressure difference. After a pressure adjustment return process where the absolute value of the detection pressure difference is less than the absolute value of the first target detection pressure, the system obtains the initial pressure response time constant, pressure repeatability boundary, pressure change rate repeatability boundary, and baseline pressure-air volume difference set.

[0008] The pressure return path is graded along the target detection pressure direction. Based on the pressure response time constant, pressure repeatability boundary, and pressure change rate repeatability boundary, the corresponding pressure regulating branch sample and return branch sample are extracted. Based on the effective supply and exhaust air volume of the two branches and the baseline same pressure air volume difference set, the same pressure different flow return path bifurcation is identified, forming two types of branch sample sets and splice branch status.

[0009] Based on the two types of branch sample sets, establish unopened branch models and open return branch models, and maintain or reset the splice branch state according to the bifurcation.

[0010] The multi-objective particle swarm optimization algorithm is used to determine the corresponding pressure control parameters based on the two-branch model;

[0011] Select the corresponding branch model and pressure control parameters according to the condition of the splice branch, determine the predicted leakage flow, and generate the supply and exhaust air execution command based on the predicted leakage flow and the deviation between the target detection pressure and the detection pressure difference.

[0012] The stability of the target detection pressure is determined based on the initial pressure response time constant, pressure repeatability boundary, and pressure change rate repeatability boundary.

[0013] Preferably, obtaining the initial pressure response time constant, pressure repeatability boundary, pressure change rate repeatability boundary, and baseline pressure-volume difference set includes:

[0014] The difference between the internal pressure of the building and the external reference pressure is defined as the detection pressure difference;

[0015] Control the supply and exhaust air actuators to change the detection pressure difference along the direction of the target detection pressure until its absolute value is less than the initial pressure of the first target detection pressure, and then return under control from the initial pressure;

[0016] The time required for the absolute value of the detected pressure difference to first reach 63.2% of the steady-state pressure increment during the pressure regulation process is defined as the initial pressure response time constant.

[0017] Based on the median absolute deviation of the detected pressure difference and pressure change rate during the pressure holding phase of the pressure regulation process, the pressure repeatability boundary and the pressure change rate repeatability boundary are determined respectively.

[0018] Based on the pressure repeatability boundary, samples in the same pressure range during the pressure regulation and return processes are paired, and the absolute values ​​of the difference between the effective supply and exhaust air volumes of the paired samples are used to form a baseline same-pressure air volume difference set.

[0019] Preferably, the extraction of voltage regulating branch samples and return branch samples includes:

[0020] Determine at least two comparison pressures from the non-final target pressures, and set a target pressure with a larger absolute value along the target detection pressure direction after each comparison pressure;

[0021] When the detected differential pressure changes along the direction of the target detected pressure and enters the pressure range centered on the comparison pressure and with the pressure repeatability boundary as half the width, and the absolute value of the pressure change rate is not greater than the pressure change rate repeatability boundary, a pressure regulating sample window with a duration not less than the initial pressure response time constant is extracted, and the median value of the absolute value of the effective supply and exhaust air volume within the pressure regulating sample window is determined as the pressure regulating branch air volume.

[0022] After the detected pressure difference continues to change until the absolute value is greater than the absolute value of the comparison pressure, the return is controlled. When the detected pressure difference re-enters the pressure range and meets the same conditions, the return sample window is extracted. The median value of the absolute value of the effective supply and exhaust air volume within the return sample window is determined as the return branch air volume.

[0023] The difference between the return branch air volume and the pressure regulating branch air volume is determined as the return bifurcation amount of the same pressure but different flow.

[0024] Preferably, the identification of bifurcation of return flows with the same pressure but different flow rates includes:

[0025] The sum of the absolute deviations of the median of the set of baseline pressure-volume differences and three times the median corrected by the normal consistency coefficient is determined as the bifurcation decision boundary, wherein the normal consistency coefficient is 1.4826.

[0026] Initialize the splice branch status to the unopened branch status;

[0027] When the absolute value of the bifurcation amount of the same pressure and different flow return path is greater than the bifurcation judgment boundary, and the absolute value of the effective supply and exhaust air volume at each sampling time in the return sample window is consistent with the deviation sign of the air volume of the pressure regulating branch, it is determined that there is a bifurcation of the same pressure and different flow return path.

[0028] The first determination will form a set of undiscovered branch sample samples based on at least two comparative pressures before the bifurcation of the same pressure and different flow return path;

[0029] After the first determination of the bifurcation of the return flow with the same pressure and different flow, at least two stable return flow samples corresponding to the comparative pressure are used to form the sample set of the return flow branch after the opening, and the splice branch state is switched to the state of the return flow branch after the opening.

[0030] Preferably, establishing models of unopened branch lines and return branch lines after opening includes:

[0031] When the absolute value of the pressure change rate is not greater than the repeatability boundary of the pressure change rate, the absolute value of the effective supply and exhaust air volume is determined as the observed value of the equivalent leakage flow rate.

[0032] Logarithmic linear fitting was performed on the absolute values ​​of the detected pressure difference and the absolute values ​​of the effective supply and exhaust air volume in the sample sets of unopened branch lines and the sample sets of open return branch lines, respectively, to obtain the leakage coefficient and pressure index of the corresponding branch lines.

[0033] A model of the unopened branch is established based on the leakage coefficient and pressure index corresponding to the unopened branch; a model of the opened return branch is established based on the leakage coefficient and pressure index corresponding to the opened return branch.

[0034] The equivalent leakage flow rate determined based on the current branch model and the absolute value of the detected pressure difference will be used as the predicted leakage flow rate.

[0035] Obtain the effective air volume and absolute air pressure inside the modular building. Based on the air mass conservation relationship and the ideal gas state relationship, establish a pressure dynamic model relating the rate of change of the absolute value of the detected pressure difference to the absolute value of the effective supply and exhaust air volume, the predicted leakage flow rate, the absolute air pressure, and the effective air volume inside the building.

[0036] Preferably, the joint branch status is maintained or reset according to the bifurcation of the same pressure and different flow return path, including:

[0037] When the splice branch state is switched to the open return branch state, the open return branch state is maintained.

[0038] During subsequent pressure regulation and return processes, continue to construct same-pressure sample pairs. When the duration of the return sample window is not less than the initial pressure response time constant, and the absolute value of the same-pressure different-flow return bifurcation amount corresponding to the return sample window is not greater than the bifurcation judgment boundary, reset the splice branch state to the unopened branch state; if the reset condition is not met, continue to maintain the open return branch state.

[0039] Preferably, a multi-objective particle swarm optimization algorithm is used to determine the corresponding pressure control parameters based on a two-branch model, including:

[0040] Multi-objective particle swarm optimization algorithm was executed for the un-uncovered branch model and the return branch model after the branch was uncovered, respectively, with the proportional control coefficient, integral control coefficient and upper limit of target pressure change rate as particle position parameters.

[0041] Substitute the particle position parameters into the pressure dynamic model of the corresponding branch model, and iteratively update the particle velocity and particle position based on the particle history preferred position and the guiding particle position.

[0042] The maximum difference between the absolute value of the detected pressure difference output by the pressure dynamic model and the absolute value of the target detected pressure is defined as the pressure overshoot. The time required for the detected pressure difference to first enter the pressure range centered on the target detected pressure and with the pressure repeatability boundary as half a width, and to remain there for no less than the initial pressure response time constant, is defined as the pressure stabilization time. The cumulative value of the absolute value of the change in the target effective supply and exhaust air volume obtained by the inverse solution of the pressure dynamic model during the pressure stabilization time is defined as the intensity of the change in the control quantity.

[0043] With the optimization objectives of reducing pressure overshoot, stabilization time, and intensity of control quantity changes, each optimization objective is normalized, and the particle position parameter closest to the ideal target point in the normalized target space is selected to form the pressure control parameters of the unopened branch and the pressure control parameters of the return branch after opening.

[0044] Preferably, the corresponding branch model and pressure control parameters are selected according to the condition of the joint branch, the predicted leakage flow rate is determined, and the supply and exhaust air execution command is generated based on the predicted leakage flow rate and the deviation between the target detection pressure and the detection pressure difference, including:

[0045] Select the corresponding branch model and pressure control parameters based on the condition of the joint branch;

[0046] Based on the target detection pressure, the current target trajectory pressure, and the selected upper limit of the target pressure change rate, a target pressure trajectory that changes along the target detection pressure direction is generated every control cycle.

[0047] The detection pressure difference deviation is determined based on the target pressure trajectory and the detection pressure difference, and the predicted leakage flow rate is determined based on the selected branch model and the absolute value of the detection pressure difference.

[0048] Based on the pressure dynamic model, predicted leakage flow, rate of change of target pressure trajectory, detection differential pressure deviation, proportional control coefficient, and integral of detection differential pressure deviation, the target effective supply and exhaust air volume is obtained by inverse solution.

[0049] Based on the pre-defined correspondence between the target effective supply and exhaust air volume and the fan speed or electric damper opening, the target effective supply and exhaust air volume is converted into supply and exhaust air execution commands;

[0050] When the supply and exhaust ventilation actuator reaches the execution limit, and the integral increment determined based on the detected pressure difference deviation causes the target effective supply and exhaust ventilation volume to continue to exceed the corresponding execution limit, the accumulation of integral quantity is suspended. The accumulation of integral quantity is resumed after the supply and exhaust ventilation actuator moves away from the execution limit.

[0051] Preferably, determining whether the target detection pressure is stably established based on the initial pressure response time constant, pressure repeatability boundary, and pressure change rate repeatability boundary includes:

[0052] The detected pressure difference is collected during the pressure closed-loop control process. The deviation of the detected pressure difference is determined based on the target pressure trajectory and the detected pressure difference, and the pressure change rate of the detected pressure difference is determined.

[0053] When the target pressure trajectory reaches the target detection pressure, the absolute value of the detection pressure difference deviation is not greater than the pressure repeatability boundary, the absolute value of the pressure change rate is not greater than the pressure change rate repeatability boundary, and the duration of the above state is not less than the initial pressure response time constant, the target detection pressure is determined to be stable.

[0054] Once the current target detection pressure is established stably, the system moves to the next target detection pressure, maintains or resets the joint branch status, and continues to generate air supply and exhaust execution commands based on the next target detection pressure until the control of each target detection pressure is completed.

[0055] Compared with the prior art, the beneficial effects of the present invention are:

[0056] 1. This invention compares the airflow of the pressure regulating branch and the return branch under the same detection pressure difference during the pressure regulation and return process of modular building air tightness testing. It also identifies the bifurcation of the same pressure and different flow return based on the baseline airflow difference set formed during the initial low-pressure pressure regulation and return process. This allows it to determine whether an equivalent leakage state change occurs after the module joint crosses the critical pressure. Compared to methods that only rely on the current detection pressure difference for control, this invention can distinguish between normal measurement fluctuations and the return flow deviation generated after the module joint is opened, avoiding misjudging abnormal pressure response caused by leakage state changes as insufficient air supply and exhaust capacity, thus improving the accuracy of joint state identification and pressure control.

[0057] 2. This invention establishes pressure-leakage flow models for both unexposed branches and returned branches after exposing them, and uses a multi-objective particle swarm optimization algorithm to determine the corresponding proportional control coefficient, integral control coefficient, and upper limit of the target pressure change rate. During the formal testing process, the appropriate model and pressure control parameters are selected based on the current state of the joint branch, and the supply and exhaust air execution commands are generated by combining the predicted leakage flow and the detection pressure difference deviation. This can reduce pressure overshoot, repeated drops, and integral accumulation caused by mismatch of a single pressure-flow model, shorten the stabilization process of the target detection pressure, reduce the reciprocating changes in fan speed or electric valve opening, and improve the stability, adaptability, and reliability of the test data in the modular building airtightness testing process. Attached Figure Description

[0058] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0059] Figure 1Functional block diagram of the modular building airtightness detection adaptive pressure closed-loop intelligent control system provided by the present invention;

[0060] Figure 2 This is a schematic diagram of the same pressure and different flow return bifurcation of the module splice provided by the present invention. Detailed Implementation

[0061] Example 1

[0062] like Figure 1 As shown, this embodiment provides a modular building airtightness detection adaptive pressure closed-loop intelligent control system, including a pressure acquisition unit, an air volume acquisition unit, an air supply and exhaust actuator, an industrial controller, and an initial dynamic baseline establishment module, a same pressure sample pairing and bifurcation identification module, a dual-branch model establishment module, a swarm intelligent parameter optimization module, a bifurcation state constraint closed-loop control module, and a pressure stability determination module, all installed in the industrial controller.

[0063] In this embodiment, modular building refers to a building space assembled from two or more building modules through horizontal joints, vertical joints, corner joints, or module connection nodes. During the testing process, the internal doors or connecting passages between the building modules for air communication remain open, so that the building modules form a connected space that can adjust the overall testing pressure difference through air supply and exhaust.

[0064] In both positive and negative pressure testing, the pressure difference is determined by subtracting the external reference pressure from the internal building pressure. During the matching of samples with the same pressure, the fitting of branch models, and the optimization of pressure control parameters, the absolute value of the pressure difference is used, and the absolute value of the flow rate after establishing a unified direction based on the target testing pressure is used for the effective supply and exhaust air volume. When outputting control commands to the supply and exhaust air actuators, the supply air direction or exhaust air direction is determined according to the positive or negative direction of the target testing pressure.

[0065] Pressure acquisition unit;

[0066] Specifically, the pressure acquisition unit includes an internal building pressure acquisition terminal and an external reference pressure acquisition terminal;

[0067] The internal pressure acquisition terminal is located inside the modular building in a position not directly impacted by the fan jet; the external reference pressure acquisition terminal is located outside the modular building, and the influence of local wind pressure on the external reference pressure is reduced by wind shields, static pressure taps or multi-directional pressure taps.

[0068] The pressure acquisition unit synchronously acquires the internal pressure of the building. and external reference pressure The two signals are then transmitted to the industrial controller, which determines the detected differential pressure using the following formula:

[0069]

[0070] In the formula: for Constant building internal pressure; for External reference pressure at any given moment; for The pressure difference detected at any given moment;

[0071] In this embodiment, the pressure acquisition unit measures from -100 Pa to +100 Pa, and the sampling period is... for Seconds; In practical applications, the measurement range of the pressure acquisition unit should cover the entire target detection pressure and the pressure overshoot allowance range;

[0072] When a modular building contains multiple internally connected areas, internal pressure acquisition terminals are set up in different building modules, and the median value of the pressure values ​​of multiple internal pressure acquisition terminals is taken as the internal pressure of the building, so as to reduce the influence of local air jet or local pressure pulsation on the detected pressure difference.

[0073] Air volume acquisition unit;

[0074] Specifically, in this embodiment, the effective supply and exhaust air volume refers to the net air volume flow rate that actually participates in the process of establishing or reducing the internal pressure difference of the modular building; during positive pressure testing, the net air flow rate supplied into the modular building is the effective supply and exhaust air volume; during negative pressure testing, the net air flow rate discharged from the modular building is the effective supply and exhaust air volume.

[0075] The air volume acquisition unit is installed in the air duct between the supply and exhaust air actuators and the modular building to collect the actual effective supply and exhaust air volume. ;

[0076] When the air volume acquisition unit directly outputs the actual volumetric flow rate of the air inside the building, the actual volumetric flow rate is directly used for baseline calculation, branch model fitting and dynamic pressure calculation.

[0077] When the air volume acquisition unit outputs standard state volumetric flow rate When the volumetric flow rate is converted to the internal air volume of the building using the following formula, it can be calculated. :

[0078]

[0079] In the formula: The standard absolute pressure used by the air volume acquisition unit; Standard absolute temperature; To measure the absolute air pressure at the site; The absolute temperature of the air inside a modular building;

[0080] The initial dynamic baseline establishment, pressure sample pairing, dual-branch model fitting, particle swarm simulation, and formal closed-loop control all use the effective supply and exhaust air volume under the same air state reference.

[0081] Ventilation supply and exhaust actuators;

[0082] Specifically, the air supply and exhaust actuator in this embodiment includes a bidirectional variable frequency fan that can be completely stopped and a connecting duct; the bidirectional variable frequency fan can switch between air supply and exhaust states, and stops air supply and exhaust when the control command is zero;

[0083] In another embodiment, positive pressure detection and negative pressure detection are respectively equipped with supply fans and exhaust fans, and the industrial controller selects the corresponding fan according to the direction of the target detection pressure.

[0084] In another embodiment, the air supply and exhaust actuator includes a constant speed fan and an electric air valve, and the actual effective air supply and exhaust volume is adjusted by the electric air valve.

[0085] The supply and exhaust air actuator has a pre-calibrated correspondence between the target effective supply and exhaust air volume and the fan frequency or the opening degree of the electric damper; during calibration, the fan is sequentially controlled at multiple frequency or valve opening calibration points, and the control command remains unchanged at each calibration point. After the air volume stabilizes, the effective supply and exhaust air volume is recorded to form a set of calibration points.

[0086]

[0087] In the formula: For the first The fan frequency or electric damper opening corresponding to each calibration point; For the first The stable and effective supply and exhaust air volume corresponding to each calibration point; This refers to the number of calibration points;

[0088] When the target effective supply and exhaust air volume Located in adjacent calibration flow and When in between, linear interpolation is used to determine the supply and exhaust air execution commands according to the following formula. :

[0089]

[0090] The flow rate is calibrated separately for positive pressure air supply and negative pressure air exhaust, and the unverified flow rate-execution command relationship is not directly shared.

[0091] Industrial controller and module connection relationships;

[0092] Specifically, the pressure acquisition unit and the air volume acquisition unit are connected to the data input terminal of the industrial controller, and the control output terminal of the industrial controller is connected to the air supply and exhaust actuator.

[0093] The initial dynamic baseline establishment module receives the detected pressure difference and effective supply and exhaust air volume, and outputs the initial pressure response time constant, pressure repeatability boundary, pressure change rate repeatability boundary, and the set of pressure and air volume differences between the baseline and the target pressure.

[0094] The pressure sample pairing and bifurcation identification module receives the initial pressure response time constant, pressure repeatability boundary, pressure change rate repeatability boundary, baseline pressure air volume difference set, as well as the detected pressure difference and effective supply and exhaust air volume during the graded pressure regulation process, and outputs the bifurcation judgment boundary, the unopened branch sample set, the return branch sample set after opening, and the splice branch status.

[0095] The dual-branch model building module receives the sample set of unexposed branches and the sample set of exposed return branches, and outputs the unexposed branch model, the exposed return branch model, and the splice branch status after holding or resetting judgment.

[0096] The swarm intelligence parameter optimization module receives the unopened branch model and the opened return branch model, and outputs the pressure control parameters of the unopened branch and the pressure control parameters of the opened return branch, respectively.

[0097] The bifurcation state constraint closed-loop control module receives the splice branch status, the corresponding branch model, the corresponding pressure control parameters, the detection pressure difference, and the target detection pressure, and outputs the air supply and exhaust execution command.

[0098] The pressure stability determination module receives the target pressure trajectory and the detected pressure difference, and outputs the determination result of whether the target detected pressure has been stably established.

[0099] System operation phase;

[0100] Specifically, the operation process of this system includes the initial dynamic baseline establishment stage, the branch pre-identification and model establishment stage, and the formal pressure closed-loop control stage in sequence.

[0101] During the initial dynamic baseline establishment phase, the supply and exhaust air actuators complete the initial pressure regulation and return cycle within the range where the absolute value of the detected pressure difference is less than the absolute value of the first target detected pressure. Based on this, the initial dynamic baseline establishment module obtains the initial pressure response time constant, pressure repeatability boundary, pressure change rate repeatability boundary, and the set of baseline pressure-airflow difference.

[0102] In this embodiment, the comparison pressure refers to the pressure value used to pair the pressure regulation branch sample and the return branch sample with the same pressure. After each comparison pressure used for branch pre-identification, a target pressure with a larger absolute value along the target detection pressure direction is set to form a return process after passing the comparison pressure.

[0103] During the branch pre-identification and model establishment stage, the supply and exhaust ventilation actuators follow the pre-identified pressure sequence, causing the detection pressure difference to pass through at least two comparative pressures along the target detection pressure direction, and then continue to reach the target pressure with a larger absolute value before being controlled to return.

[0104] The pressure regulating branch samples and return branch samples collected during the branch pre-identification and model building stage are used for same-pressure different-flow return bifurcation identification, branch model identification and pressure control parameter optimization, and are not used as formal airtightness test result data.

[0105] When the same pressure sample pairing and bifurcation identification module identifies the same pressure different flow return bifurcation, and both the unexposed branch sample set and the exposed return branch sample set contain at least two stable samples corresponding to different comparison pressures, the dual branch model building module builds the unexposed branch model and the exposed return branch model respectively, and the swarm intelligence parameter optimization module obtains the corresponding pressure control parameters respectively.

[0106] After completing branch pre-identification, model establishment, and parameter optimization, the supply and exhaust ventilation actuators stop actively adjusting the pressure, causing the detected differential pressure to fall back towards zero. When the following conditions are met continuously for a period of not less than the initial pressure response time constant, it is determined that the initial pressure conditions for formal pressure closed-loop control have been re-established:

[0107]

[0108]

[0109] In the formula: This represents the pressure repeatability boundary. This represents the repeatability boundary of the pressure change rate;

[0110] After re-establishing the initial pressure conditions, the pressure error integral state is cleared to zero, the target pressure trajectory cache of the branch pre-identification stage is cleared, and the established branch model, corresponding pressure control parameters, bifurcation judgment boundary and current splice branch state are retained, and the formal pressure closed-loop control stage is entered.

[0111] Example 2

[0112] This embodiment uses a modular building formed by connecting four building modules as an example; the effective air volume inside the modular building... for cubic meters, measuring the absolute pressure of air at the testing site. for Pa;

[0113] The positive pressure target detection pressures set in this embodiment are as follows: Pa, Pa, Pa, Pahe Pa;

[0114] Before performing initial dynamic baseline establishment and branch pre-identification, record the building's internal air temperature, external air temperature, absolute air pressure, and external natural pressure difference;

[0115] When the ventilation actuator is not activated, the internal pressure of the building and the external reference pressure are continuously collected. When the amplitude of the natural pressure difference fluctuation exceeds the absolute value of the first target detection pressure, or when the natural pressure difference continues to drift in one direction and the pressure change rate exceeds the initial pressure change rate noise boundary, the branch pre-identification is suspended and resumed after the external pressure disturbance is reduced.

[0116] Determine the initial pressure response time constant;

[0117] Specifically, the initial pressure response time constant in this embodiment is the equivalent response characteristic time defined by the time required for the detected pressure difference to first reach 63.2% of the steady-state pressure increment. It is used to determine the duration of the sample window and the duration of pressure stabilization, and does not indicate that the modular building strictly belongs to a first-order linear system.

[0118] The initial dynamic baseline establishment module first initializes the splice branch status to the unopened branch status;

[0119] Before the initial pressure regulation begins, keep the effective supply and exhaust air volume of the supply and exhaust air actuator at zero, continuously collect no less than fifty differential pressure samples, and calculate the pressure change rate sample under zero pressure regulation state based on adjacent differential pressure samples.

[0120] No. The rate of change of pressure at each sampling time is determined by the following formula:

[0121]

[0122] For any sample set The absolute deviation of the median is determined by the following formula:

[0123]

[0124] The median absolute deviation of the pressure change rate samples under zero pressure regulation conditions, corrected by the normality coefficient, is multiplied by three to determine the initial pressure change rate noise boundary. :

[0125]

[0126] In this embodiment, the baseline pressure air volume difference is approximately normally distributed, and the consistency correction coefficient is 1.4826. When other data distribution assumptions are adopted or the correction relationship is determined by calibration, a consistency correction coefficient that matches the corresponding data distribution can be used.

[0127] The control mechanism of the supply and exhaust air actuators changes the detected pressure difference from near zero Pa along the positive pressure direction to... Pa, and keep the initial voltage regulation command unchanged; due to The absolute value of Pa is less than the first target detection pressure. The absolute value of Pa is used to determine the initial pressure response time constant;

[0128] Five-point sliding median filtering is applied to the collected differential pressure to reduce the impact of single-point impulse noise on the first arrival time;

[0129] When the median of the absolute value of the pressure change rate within each of three consecutive pressure judgment windows is not greater than the initial pressure change rate noise boundary, the median value of the detected pressure difference within the third pressure judgment window is determined as the steady-state detected pressure difference. Each pressure determination window includes twenty consecutive sampling points.

[0130] The initial voltage regulation start time is denoted as... The median value of the differential pressure sample measured before the pressure regulation begins is recorded as... Then the steady-state pressure increment for:

[0131]

[0132] In the filtered main response segment of the detected differential pressure, determine the moment when the detected differential pressure first satisfies the following equation. :

[0133]

[0134] The initial pressure response time constant Determined as:

[0135]

[0136] The maximum holding time for the initial pressure regulation command is 30 seconds. If a steady-state detection pressure difference that meets the above conditions is not formed within 30 seconds, or if the detection pressure difference does not reach 63.2% of the steady-state pressure increment, the initial dynamic baseline establishment is deemed invalid, and the parameter calculation is stopped. After checking the supply and exhaust ventilation actuators, pressure acquisition units, and building opening status, the initial pressure regulation process is re-executed.

[0137] In this exemplary configuration, the following is obtained: .

[0138] Determine the pressure repeatability boundary and the pressure change rate repeatability boundary;

[0139] Specifically, during the initial pressure adjustment and holding phase, a sample set of differential pressure is obtained. and pressure change rate sample set ;

[0140] Pressure repeatability boundary Determined according to the following formula:

[0141]

[0142] Pressure change rate repeatability boundary Determined according to the following formula:

[0143]

[0144] In this exemplary configuration, we get: ; .

[0145] Form a baseline set of air volume differences at the same pressure;

[0146] Specifically, after determining the initial pressure response time constant and two repeatability boundaries, the initial dynamic baseline establishment module re-executes the graded pressure regulation and return cycle below the first target detection pressure;

[0147] This embodiment is in Pa, Pa, Pa, Pa, Pa, Pa, Pahe Set the initial comparison pressure for each Pa;

[0148] The detected pressure difference passes through each initial comparison pressure along the positive pressure direction; after reaching each initial comparison pressure, the target pressure trajectory remains unchanged; when the detected pressure difference satisfies:

[0149]

[0150]

[0151] And the duration is not less than the initial pressure response time constant. At that time, extract the corresponding voltage regulation branch stability window;

[0152] Detected pressure difference reaches After completing the hold, proceed back in reverse order, level by level. Pa, Pa, Pa, Pa, Pa, Pahe Pa; whenever an initial comparison pressure is reached, the stability window of the return branch is extracted under the same conditions;

[0153] Calculate the median absolute values ​​of effective supply and exhaust air volumes in the stable windows of the pressure regulating branch and the return branch under the same initial comparison pressure. Use the absolute value of the difference between these two values ​​as a baseline pressure-same-volume difference set. ;

[0154] The baseline pressure airflow difference set should include no less than five valid baseline pressure sample pairs; sample windows that cannot simultaneously meet the pressure range condition, pressure change rate condition, and duration condition should not be added to the baseline pressure airflow difference set.

[0155] In this exemplary configuration, the following is obtained: .

[0156] Determine the bifurcation boundary and verify the measurement error;

[0157] Specifically, the same-pressure sample pairing and bifurcation identification module uses the baseline same-pressure air volume difference set... Determine the bifurcation decision boundary :

[0158]

[0159] The median of the above baseline pressure-volume difference set is cubic meters per hour, median absolute deviation is cubic meters per hour, therefore:

[0160]

[0161] When the median absolute deviation of the baseline pressure airflow difference set is zero, the baseline pressure sample pairs corresponding to different initial comparison pressures are added and the calculation is repeated; if the median absolute deviation is still zero after adding samples, the baseline acquisition is re-performed using an airflow acquisition unit with higher resolution.

[0162] The upper limit of repeatability error given in the calibration certificate of the air volume acquisition unit, or the upper limit of the range of air volume obtained by repeated acquisitions under the same supply and exhaust air execution command, shall be determined as the upper limit of air volume measurement uncertainty. ;

[0163] The upper limit of dynamic gas storage flow rate caused by residual pressure changes within the sample window Determined according to the following formula:

[0164]

[0165] In this exemplary configuration:

[0166]

[0167] The sum of the upper limit of the air volume measurement uncertainty and the upper limit of the dynamic gas storage flow rate is determined as the upper limit of the total flow uncertainty. :

[0168]

[0169] In this exemplary configuration, for cubic meters per hour, therefore:

[0170]

[0171] The boundary for determining a bifurcation should satisfy:

[0172]

[0173] The fork determination boundary in this exemplary configuration cubic meters per hour greater than the upper limit of uncertainty of total flow rate The flow rate is cubic meters per hour, therefore the bifurcation determination boundary is valid;

[0174] when Not greater than If the baseline cannot distinguish between measurement error, residual gas flow, and branch bifurcation, the baseline sample should be re-collected or the accuracy of pressure and air volume acquisition should be improved.

[0175] Extract samples from the voltage regulating branch and the return branch;

[0176] Specifically, select from non-final target detection pressures. Pa, Pa, Pahe Pascal as a comparative pressure;

[0177] When the pressure difference is detected, it enters along the positive pressure direction to compare the pressure. A pressure interval centered at the pressure repeatability boundary with a width of half its width, satisfying:

[0178]

[0179]

[0180] At that time, the voltage regulating branch sample window is recorded.

[0181] The duration of the sample window for the pressure regulating branch must not be less than the initial pressure response time constant. The sampling period in this embodiment is: Seconds, therefore each effective sample window includes at least:

[0182] A series of sampling points;

[0183] The median absolute value of the effective supply and exhaust air volume within the sample window of the pressure regulating branch is determined as the air volume of the pressure regulating branch:

[0184]

[0185] The pressure difference was detected sequentially through Pa, Pa, Pahe Pa continued to change to Pa, then the detected pressure difference is controlled sequentially on the return journey. Pa, Pa, Pahe Pa;

[0186] When the detected pressure difference during the return process re-enters the pressure range corresponding to the comparison pressure, and meets the same pressure change rate condition and window duration condition, the median value of the absolute value of the effective supply and exhaust air volume within the sample window of the return branch is determined as the return branch air volume:

[0187]

[0188] Calculate the bifurcation of the return flow under the same pressure but different flow:

[0189]

[0190] The values ​​in the table below are example data constructed based on the two branch models described below, used to illustrate the process of pairing samples under the same pressure, calculating the bifurcation amount, and fitting the branch model.

[0191] 10Pa 53.6 m³ / h 76.6 m3 / h 23.0 m3 / h 20Pa 84.1 m³ / h 122.7 m3 / h 38.6 m3 / h 30Pa 109.5 m3 / h 161.6 m3 / h 52.1 m³ / h 40Pa 132.0 m3 / h 196.6 m3 / h 64.6 m3 / h

[0192] In this embodiment, the unopened branch state refers to the pressure-flow state formed when the detected pressure loading history has not caused a significant change in the equivalent leakage state of the modular building.

[0193] In this embodiment, the return branch state after opening refers to the pressure-flow state formed when the detected pressure difference returns to the original pressure range after the detected pressure difference exceeds the critical pressure, but the equivalent leakage state of the modular building has not yet recovered to the state before the branch is opened.

[0194] The joint branch status in this embodiment is the equivalent leakage status determined based on the overall pressure-flow response of the modular building, and does not mean that a visible opening of a specific module joint must be directly detected.

[0195] In this embodiment, the same pressure, different flow return bifurcation refers to the situation where, after the detected pressure difference passes through the comparison pressure along the direction of the target detected pressure, it continues to change to a target pressure with a larger absolute value, and then returns under control to the pressure range where the comparison pressure is located. At this point, the effective supply and exhaust air volume of the return branch produces a continuous deviation from the effective supply and exhaust air volume of the previous pressure regulating branch, exceeding the baseline fluctuation and measurement uncertainty.

[0196] Identify bifurcations of return paths with the same pressure but different flow rates and form a set of branch sample sets;

[0197] like Figure 2 As shown, a module seam is formed between module unit A and module unit B. In the unopened state, the equivalent leakage channel of the module seam is small, corresponding to the pressure regulating branch. When the detected differential pressure increases along the target detected pressure direction and exceeds the critical opening pressure of the module seam, the equivalent leakage channel of the module seam increases, and it does not immediately return to the unopened state during the return process of the detected differential pressure, corresponding to the return branch after opening. At the same comparison pressure... At the location, the absolute value of the effective supply and exhaust air volume corresponding to the return branch is higher than the absolute value of the effective supply and exhaust air volume corresponding to the pressure regulating branch. The difference between the two is the same pressure and different flow return bifurcation amount. By comparing the bifurcation amount with the bifurcation determination boundary, it is possible to determine whether the equivalent leakage state of the modular building has switched from the unopened branch state to the opened return branch state.

[0198] Specifically, when the absolute value of the bifurcation amount of the same pressure and different flow return path is greater than the bifurcation judgment boundary, and the absolute value of the effective supply and exhaust air volume at each sampling time in the return path sample window is consistent with the deviation sign of the air volume of the pressure regulating branch, it is determined that there is a same pressure and different flow return path bifurcation.

[0199] In this example, the detected pressure difference is determined by... Pa's return journey At Pa, the absolute value of the bifurcation of the same pressure and different flow return is cubic meters per hour, greater than the bifurcation determination boundary cubic meters per hour;

[0200] Meanwhile, the absolute values ​​of the effective supply and exhaust air volume at each sampling time within the sample window of the return branch are greater than the air volume of the pressure regulating branch. Therefore, it is determined that there is a return bifurcation with the same pressure but different flow, and the status of the splice branch is switched from the unopened branch status to the opened return branch status.

[0201] The set of samples of stable pressure regulating branches corresponding to at least two comparative pressures before the bifurcation of the same pressure and different flow return path will be used to form the unrevealed branch sample set:

[0202]

[0203] The set of return branch samples after the initial identification of the bifurcation of the same-pressure heterogeneous flow will be composed of at least two stable return branch samples corresponding to the comparative pressure:

[0204]

[0205] When a bifurcation of the same pressure and different flow is first identified, if the remaining comparative pressure in the current return path is insufficient to form at least two open return branch samples, the open return branch model is not established immediately.

[0206] In this case, the branch pre-identification process is repeated, and at least two intermediate comparison pressures are added below the absolute value of the initial bifurcation pressure; the intermediate comparison pressures may be within the range of the initial baseline pressure, but the return branch samples must be re-collected after the detection pressure difference has exceeded the bifurcation pressure after they are activated.

[0207] Before obtaining at least two stable return samples corresponding to different comparison pressures, the system continues to operate using the unexposed branch model and the unexposed branch pressure control parameters.

[0208] The operating mode when the return flow is bifurcised under the same pressure but different flow was not identified;

[0209] Specifically, when all the comparison pressures in the pre-identified pressure sequence have completed the pairing of pressure regulating branch samples and return branch samples, and the absolute value of the bifurcation amount of the same pressure and different flow return corresponding to each comparison pressure is not greater than the bifurcation judgment boundary, the splice branch state remains as the unopened branch state, and the set of return branch samples after opening is not forcibly generated.

[0210] At this time, the dual-branch model establishment module first establishes the unrevealed branch model based on the unrevealed branch sample set. The swarm intelligence parameter optimization module determines the unrevealed branch pressure control parameters only for the unrevealed branch model. The formal pressure closed-loop control stage uses the unrevealed branch model and the unrevealed branch pressure control parameters for operation.

[0211] When the bifurcation of the same pressure and different flow is first identified in the subsequent diagnostic return, the system continues to use the unopened branch model and the unopened branch pressure control parameters to maintain control, while collecting at least two comparison pressures corresponding to the stable return after opening.

[0212] Once the sample set of the return branch after the exposure meets the model identification requirements, the model of the return branch after exposure is established, and the pressure control parameters of the return branch after exposure are determined. The system is then switched from the single-branch operation mode without exposure to the dual-branch adaptive pressure closed-loop control mode.

[0213] Establish models for undiscovered branch routes and return branch routes after discovery;

[0214] Specifically, when the absolute value of the pressure change rate is not greater than the repeatability boundary of the pressure change rate, the absolute value of the effective supply and exhaust air volume is determined as the observed value of the equivalent leakage flow rate;

[0215] Branches are expressed by the following formula Pressure-leakage flow relationship:

[0216]

[0217] In the formula: This indicates that the branch road has not been opened. Indicates the status of the return branch after it is opened; branch road The corresponding equivalent leakage flow rate; branch road The corresponding leakage coefficient; branch road The corresponding pressure index;

[0218] Take the natural logarithm of the above expression:

[0219]

[0220] Using the unrevealed branch sample set respectively and the sample set of return branches after the opening Perform constrained least squares fitting and convert the pressure index Constraints to between;

[0221] Based on the aforementioned exemplary data, the unrevealed branch model is obtained:

[0222]

[0223] The fitting yields the return branch model after the opening is completed:

[0224]

[0225] Based on the current status of the splice branch Select the corresponding branch model and calculate the predicted leakage flow:

[0226] .

[0227] Determine the effective internal air volume;

[0228] Specifically, in this embodiment, the internal effective air volume refers to the air space volume that is connected to the internal air of the modular building during the detection process, and whose internal pressure changes together with the detection pressure difference;

[0229] When determining the effective internal air volume, the net air volume inside each building module is added together, and the rooms, corridors, open equipment shafts, ventilation ducts and ceiling-connected spaces that are connected to the air inside the building during the testing process are included in the effective internal air volume.

[0230] The volume occupied by enclosed cavities isolated from the testing space, enclosed equipment housings, unopened independent equipment wells, and solid structures is not included in the effective internal air volume.

[0231] When it cannot be directly determined whether a certain space is connected to the detection space, temporary pressure acquisition points are set up in both the space and the main detection space. During the pressure adjustment process, if the two pressure changes are in the same direction and the steady-state pressure difference is not greater than the pressure repeatability boundary, the space is included in the internal effective air volume; otherwise, it is not included.

[0232] Once the internal effective air volume is determined, it remains unchanged during the same branch model identification, parameter optimization, and formal closed-loop control process.

[0233] Establish a dynamic pressure model;

[0234] Specifically, under the condition that the absolute temperature of the air remains approximately constant within a single pressure regulating window, a pressure dynamic model is established based on the air mass conservation relationship and the ideal gas state relationship;

[0235] The absolute value of the detected pressure difference is denoted as... :

[0236]

[0237] The direction of the target detection pressure is denoted as:

[0238]

[0239] The approximate absolute pressure inside a modular building is:

[0240]

[0241] The pressure dynamic model is as follows:

[0242]

[0243] When the ratio of the absolute value of the detected pressure difference to the absolute pressure of the ambient air meets the following condition:

[0244]

[0245] At that time, adopt Approximate substitution ,get:

[0246]

[0247] The maximum detection pressure difference in this embodiment is Pascal, whose ratio to the absolute pressure of air is less than 100 kPa. Therefore, the above approximate model is adopted;

[0248] When the ratio of the absolute value of the detected pressure difference to the absolute pressure of the ambient air is greater than At that time, in the process of pressure dynamic model and target effective supply and exhaust air volume inverse solution, it is used No longer in use approximate.

[0249] The pressure control parameters are determined using a multi-objective particle swarm optimization algorithm.

[0250] Specifically, the swarm intelligence parameter optimization module executes a multi-objective particle swarm optimization algorithm for both the unrevealed branch model and the revealed return branch model.

[0251] The particle position parameters for each particle are:

[0252]

[0253] In the formula: This is the proportional control coefficient; These are integral control coefficients; The upper limit of the target pressure change rate;

[0254] Based on the initial pressure response time constant Determine the search range for particle position parameters:

[0255]

[0256]

[0257]

[0258] In the formula, The maximum pressure increment between adjacent target detection pressures, in this embodiment, is... Pa;

[0259] Particle swarm includes There are particles, and the maximum number of iterations is . Next; the initial position of the particles is generated uniformly within the search range of each parameter; the initial velocity of the particles is generated uniformly between negative 10% and positive 10% of the width of the corresponding parameter search range;

[0260] No. The particle in the first The velocity and position in the next iteration are updated according to the following formula:

[0261]

[0262]

[0263] In the formula: Preferred positions for particle history; The position of the guiding particle selected from the non-dominated solution set; The inertia coefficient; and The learning coefficient; and A random number between zero and one; the inertia coefficient is determined by... linearly reduced to Learning coefficient and All ;

[0264] When a particle exceeds the parameter search boundary after updating, the parameter is truncated to the corresponding boundary value, and the particle velocity in the dimension of that parameter is set to zero.

[0265] For each particle, the forward Euler method is used to perform discretization calculations on the pressure dynamic model:

[0266]

[0267] The simulation of each particle uses the same target pressure trajectory generation rules, target effective supply and exhaust air volume limiting rules, and integral limiting rules as the formal closed-loop control.

[0268] The initial detection pressure difference for each particle is taken as the absolute value of the detection pressure difference when control begins at the corresponding target pressure level. The initial target pressure trajectory takes the same value, and the pressure error integral state is set to zero.

[0269] Maximum simulation time for each particle simulation Determined according to the following formula:

[0270]

[0271] In the formula, The time required for the target pressure trajectory to change from the initial pressure to the target detection pressure, according to the upper limit of the target pressure change rate encoded by the particle;

[0272] The particle simulation is considered to have reached a steady state only if the following conditions are met simultaneously: the target pressure trajectory has reached the target detection pressure; the absolute value of the deviation between the detection pressure difference and the target detection pressure is not greater than the pressure repeatability boundary; the absolute value of the pressure change rate is not greater than the pressure change rate repeatability boundary; and the duration of the above state is not less than the initial pressure response time constant.

[0273] Pressure overshoot Determined according to the following formula:

[0274]

[0275] Stabilization time This is the total time elapsed from the moment the current target detection pressure command is issued until the above-mentioned stability conditions continuously satisfy the initial pressure response time constant;

[0276] Intensity of change in control quantity The cumulative change in the target effective supply and exhaust air volume is determined according to the following formula:

[0277]

[0278] In the formula, This refers to the number of control cycles from the issuance of the target detection pressure command to the attainment of a steady state; the same control cycle is used when comparing different particles. ;

[0279] The feasibility priority rule is used to compare particles: any particle that meets the stability condition takes precedence over any particle that does not meet the stability condition; when both particles meet the stability condition, the Pareto dominance relationship is judged according to the pressure overshoot, stabilization time and the intensity of the change in the control quantity; when neither particle meets the stability condition, the particle with the smaller pressure deviation at the end of the maximum simulation time is selected first.

[0280] When all three optimization objectives of a candidate particle are no worse than those of another particle, and at least one objective is better than that of the other particle, the candidate particle is determined to dominate the other particle.

[0281] When the set of non-dominated solutions exceeds the storage capacity, calculate the crowding distance of the non-dominated solutions in the normalized target space, prioritize deleting non-dominated solutions with smaller crowding distances, and randomly select guide particles from non-dominated solutions with larger crowding distances.

[0282] Each optimization objective is normalized according to the following formula:

[0283]

[0284] When the maximum and minimum values ​​of a certain optimization objective are the same, the normalized value of the optimization objective is uniformly set to zero.

[0285] Calculate the distance between the candidate particle and the ideal target point:

[0286]

[0287] For each branch model, five independent multi-objective particle swarm optimizations are performed, each using a different pseudo-random number seed. The non-dominated solution sets formed by the five optimizations are merged, and the non-dominated screening is performed again. The particle position parameter closest to the ideal target point is determined as the final pressure control parameter of the corresponding branch.

[0288] In this exemplary configuration, the pressure control parameters for the unopened branch output by the multi-objective particle swarm optimization algorithm are as follows:

[0289]

[0290] The pressure control parameters for the return branch after the circuit is opened are as follows:

[0291]

[0292] The two sets of parameters above are the algorithm outputs used to describe the control calculation process.

[0293] Re-establish the initial pressure conditions for formal control;

[0294] Specifically, after completing branch pre-identification, model fitting, and parameter optimization, the supply and exhaust ventilation actuators stop actively supplying air, allowing the detected pressure difference to fall back to zero Pa by relying on the natural leakage of the modular building.

[0295] When the absolute value of the detected differential pressure is not greater than the pressure repeatability boundary, the absolute value of the pressure change rate is not greater than the pressure change rate repeatability boundary, and the duration of the above state is not less than the initial pressure response time constant, the initial pressure conditions for formal closed-loop control are determined to be established.

[0296] Set the pressure error integral state to zero, set the target pressure trajectory to the current detected differential pressure, and restart according to... Pa, Pa, Pa, Pahe The formal target detection pressure sequence of Pa implements pressure closed-loop control.

[0297] Generate a bidirectional rate-limited target pressure trajectory;

[0298] Specifically, the absolute value of the target pressure at the current stage is denoted as... ;

[0299] When formally establishing target detection pressure:

[0300]

[0301] When performing a diagnostic return:

[0302]

[0303] Absolute value of target pressure trajectory Generate according to the following formula, controlling each cycle:

[0304]

[0305] In the formula, Indicates: When Time to take ,when Time to take ,when Time to take ;

[0306] The target pressure trajectory formula can both increase the absolute value of the target pressure according to the upper limit of the rate and decrease the absolute value of the target pressure according to the upper limit of the rate during the diagnostic return process.

[0307] The target pressure trajectory is:

[0308]

[0309] In the formula:

[0310] .

[0311] Execute closed-loop control with bifurcation state constraints;

[0312] Specifically, the absolute value of the detected pressure difference is recorded as:

[0313]

[0314] The detected differential pressure deviation is:

[0315]

[0316] The points status is updated according to the following formula:

[0317]

[0318] Determine the predicted leakage flow rate based on the current branch model:

[0319]

[0320] Convert the predicted leakage flow rate to cubic meters per second:

[0321]

[0322] The rate of change of the target pressure trajectory is:

[0323]

[0324] Based on the pressure dynamic model, the target effective supply and exhaust air volume without amplitude limitation is solved:

[0325]

[0326] This embodiment uses a supply and exhaust air actuator that can be completely stopped, with a minimum target effective supply and exhaust air volume of zero;

[0327] The maximum controllable flow rate of the supply and exhaust ventilation actuator is denoted as... The actual effective supply and exhaust air volume of the target issued shall be determined according to the following formula:

[0328]

[0329] when At that time, When set to zero, the supply and exhaust actuators are not driven to actively adjust the pressure in the opposite direction. Instead, the natural leakage of the modular building is used to reduce the absolute value of the detected pressure difference.

[0330] when When the condition is positive, the air supply and exhaust actuators perform air supply actions that increase the pressure inside the building; when... When the value is negative, the ventilation actuator performs ventilation to reduce the pressure inside the building;

[0331] Based on the pre-defined relationship between the target effective supply and exhaust air volume and the fan frequency or electric damper opening degree under the corresponding working direction, Convert to supply and exhaust ventilation execution commands;

[0332] When the splice branch is in the unopened branch state, the unopened branch model and unopened branch pressure control parameters are selected; when the splice branch is in the opened return branch state, the opened return branch model and opened return branch pressure control parameters are selected.

[0333] For those that cannot be completely stopped and have a minimum stable operating flow rate The ventilation supply and exhaust actuator, when Stop running when; Increase from zero and reach Starts at a specific time; after starting, it only runs when... Stop when the flow rate drops below a pre-defined shutdown threshold to avoid frequent start-ups and shutdowns near the minimum stable operating flow rate.

[0334] Points limit;

[0335] Specifically, when And detect pressure difference deviation When the time is positive, pause the points status update:

[0336]

[0337] when And detect pressure difference deviation When the score is negative, the points status update will also be paused.

[0338] When the effective supply and exhaust air volume of the target without limit returns to zero... After that, the points status will be updated again.

[0339] The target detection pressure has been stabilized.

[0340] Specifically, the pressure stability determination module continuously collects the target pressure trajectory and detects the pressure difference;

[0341] When the target pressure trajectory has reached the current target detection pressure, and simultaneously the following conditions are met:

[0342]

[0343]

[0344] Furthermore, the duration of the aforementioned state is not less than the initial pressure response time constant. At that time, it is determined that the current target detection pressure has been stably established;

[0345] In this exemplary configuration, the stability conditions are:

[0346]

[0347]

[0348] The continuous holding time is not less than Second.

[0349] Perform a diagnostic return and maintain or reset the joint branch status;

[0350] Specifically, when the splice branch status is switched to the open return branch status, it will not be directly reset due to the decrease in the absolute value of the detection pressure difference;

[0351] During the formal pressure closed-loop control phase, when the current target detection pressure is stably established and the splice branch is in the state of the open return branch, a diagnostic return is performed before entering the next target detection pressure with a larger absolute value.

[0352] When there is no formal comparison pressure with a smaller absolute value for the current target detection pressure, the reset judgment is not performed at this pressure level. The current splice branch state is maintained, and the diagnostic return is performed after the next target detection pressure is stably established.

[0353] When a comparison pressure with a smaller absolute value exists, the comparison pressure closest to the absolute value of the current target detection pressure is selected as the reset comparison pressure from among the comparison pressures. ;

[0354] Set the current target value of the target pressure trajectory to:

[0355]

[0356] The diagnostic return target pressure trajectory is generated using the aforementioned two-way rate-limited formula.

[0357] When the detected pressure difference meets:

[0358]

[0359]

[0360] And when the duration is not less than the initial pressure response time constant, extract the sample window of the reset return branch;

[0361] The median value of the absolute values ​​of the effective supply and exhaust air volumes within the sample window of the reset return branch is determined as the reset return branch air volume. ;

[0362] Extract the sample window of the stable pressure regulating branch when the target detected pressure direction passes through the reset comparison pressure from the current formal pressure closed-loop control cycle, and determine the median value of the absolute value of its effective supply and exhaust air volume as the air volume of the reset pressure regulating branch. ;

[0363] When no valid pressure regulating branch sample window corresponding to the reset comparison pressure is formed in the current formal control cycle, this diagnostic return is only used for pressure return and the joint branch status reset judgment is not performed.

[0364] Calculate the resetting pressure-same-pressure-different-flow return bifurcation amount:

[0365]

[0366] When the following conditions are met:

[0367]

[0368] When the condition is met, the splice branch status will be reset to the unopened branch status; if the condition is not met, the return branch status after opening will continue to be maintained.

[0369] After completing the reset judgment, the absolute value of the target pressure for the current stage is set. Set the absolute value of the detection pressure for the next target, and select the corresponding branch model and pressure control parameters based on the state of the splice branch after the reset judgment;

[0370] When the current target detection pressure is the last target detection pressure in the formal target pressure sequence, a diagnostic backtracking can be performed after stable data acquisition is completed and the branch reset result is recorded, without proceeding to the next target detection pressure.

[0371] Example 3

[0372] In this embodiment, the negative pressure target detection pressures are sequentially negative. Pa, negative Pa, negative Pa, negative Pa and negative Pa;

[0373] The supply and exhaust ventilation actuators switch to exhaust mode to exhaust air from inside the modular building; the detection pressure difference is still determined by subtracting the external reference pressure from the internal pressure of the building, so the detection pressure difference under negative pressure detection conditions is a negative value.

[0374] The positive pressure test condition and the negative pressure test condition respectively perform the initial dynamic baseline establishment, branch pre-identification, branch model identification and pressure control parameter optimization;

[0375] Positive pressure detection conditions generate sets of positive pressure baselines and pressure airflow differences. Positive pressure bifurcation determination boundary Positive pressure unopened branch model Model of return branch after positive pressure is turned on And the corresponding positive pressure control parameters;

[0376] Negative pressure detection conditions generate negative pressure baselines and pressure airflow difference sets respectively. Negative pressure bifurcation determination boundary Negative pressure unopened branch model Model of return branch after negative pressure is turned on And the corresponding negative pressure control parameters;

[0377] Positive pressure detection and negative pressure detection are respectively set to the joint branch status. and When switching from positive pressure testing to negative pressure testing, the joint branch state at the end of the positive pressure testing is not directly used; the initial dynamic baseline of negative pressure is re-established before the start of negative pressure testing, and... Initialize to a negative pressure, unopened branch state;

[0378] During initial negative pressure adjustment, control the detected pressure difference to change along the negative pressure direction until it reaches negative. Pa, and in accordance with the negative Pa to negative The initial comparison pressure sequence of Pa performs graded hold and return, forming a negative pressure baseline and pressure air volume difference set;

[0379] Select negative pressure from the non-final stage negative pressure target detection pressure. Pa, negative Pa, negative Pa and negative Pascal as a comparative pressure;

[0380] The detected pressure difference changes sequentially from negative pressure to negative pressure. Pa, negative Pa, negative Pa, negative Pa and negative Pa, then by the negative The journey back to Pa was under control, and it passed through the negative area again. Pa, negative Pa, negative Pa and negative Pa;

[0381] For any negative pressure comparison pressure The following formula is used to determine whether the detected pressure difference has entered the corresponding pressure range:

[0382]

[0383] The air volume of the pressure regulating branch and the air volume of the return branch are both taken as the median value of the absolute value of the exhaust volume;

[0384] The bifurcation amount of the return flow under the same pressure but different flow is still determined according to the following formula:

[0385]

[0386] When the absolute value of the bifurcation exceeds the negative pressure bifurcation judgment boundary, and the absolute value of the exhaust volume at each sampling time in the return branch sample window is consistent with the deviation sign of the air volume of the pressure regulating branch, it is determined that there is a return bifurcation with the same pressure and different flow during the negative pressure detection process.

[0387] The negative pressure branch model is also fitted using the absolute value of the detected pressure difference and the absolute value of the effective supply and exhaust air volume;

[0388] In negative pressure closed-loop control:

[0389]

[0390] The target effective supply and exhaust air volume amplitude is obtained by inverse solving the same pressure dynamic model as positive pressure detection. The supply and exhaust air actuator is based on... The execution command converts the target effective supply and exhaust air volume into the exhaust direction;

[0391] Without repeated testing and confirmation under both positive and negative pressure conditions, leakage coefficients, pressure indices, bifurcation judgment boundaries, and pressure control parameters are not shared between positive and negative pressure testing conditions.

[0392] Example 4

[0393] To illustrate this with a set of specific examples, the effective air volume inside a modular building is... cubic meters; absolute air pressure is Pa; control cycle is seconds; target detection pressure is Pa; the maximum effective supply and exhaust air volume of the supply and exhaust actuator is cubic meters per hour; initial detected pressure difference is zero Pa; initial target pressure trajectory is zero Pa; initial pressure error integral state is zero; random pressure noise and random air volume noise are not superimposed during the process;

[0394] The controlled object initially adopts an unrevealed branch model:

[0395]

[0396] When the absolute value of the detected pressure difference first reaches At time 8, the controlled object is switched to the backhaul branch model after being opened:

[0397]

[0398] Simulation time reached At 10:00, the controlled object is restored to the unexposed branch model to simulate the delayed recovery of the splice seam;

[0399] The fixed single-branch control method consistently uses the unopened branch model and unopened branch pressure control parameters throughout the entire process.

[0400]

[0401] The dual-branch control method of this invention employs an unopened branch model and unopened branch pressure control parameters when the controlled object is in the unopened branch state, and employs an opened return branch model and opened return branch pressure control parameters when the controlled object is in the opened return branch state.

[0402]

[0403] Both control methods employ the aforementioned bidirectional rate-limited target pressure trajectory, target effective supply and exhaust air volume zero-to-maximum flow rate limit, and integral limit rules;

[0404] Based on the aforementioned discrete pressure dynamic model, cycle-by-cycle calculations yielded the following results:

[0405] Fixed single-branch control 19.67 Pa 78.5s 313.0 m3 / h This invention features dual-branch control. 0.11Pa 52.4s 306.6 m3 / h

[0406] The above results correspond to the branch model, branch switching conditions, recovery time, execution limits and control parameters explicitly defined in this embodiment;

[0407] Under the conditions set in this embodiment, the fixed single-branch control method results in a large pressure overshoot after the controlled object returns to the unopened branch state due to the accumulation of previous model mismatch and integral state. The dual-branch control method of the present invention selects the corresponding model and pressure control parameters according to the splice branch state, thereby reducing the impact of the model mismatch on the pressure closed-loop process.

[0408] Example 5

[0409] When conducting actual equipment or engineering verification, under the conditions of the same modular building, the same closed state of doors and windows, the same internal connection state, and the same air supply and exhaust actuator, fixed single-branch control and the dual-branch control of the present invention are respectively executed;

[0410] Each control method is independently repeated no less than three times; before each test, the test pressure difference is restored to near zero Pa, and the assembly state of the module joints, the state of the sealing material, and the state of the openings inside the building are kept consistent.

[0411] The following data were recorded for each actual verification: detection time; internal building pressure; external reference pressure; detection pressure difference; target pressure trajectory; effective supply and exhaust air volume; supply and exhaust air execution command; joint branch status; pressure regulating branch sample window; return branch sample window; same pressure different flow return bifurcation amount; pressure overshoot; pressure stabilization time; cumulative change in target effective supply and exhaust air volume.

[0412] The actual verification report also records the model, range, accuracy, sampling cycle, effective air volume inside the modular building, on-site air temperature, absolute air pressure, and external natural pressure difference of the pressure acquisition unit and air volume acquisition unit;

[0413] In actual verification, the pressure overshoot is determined according to the following formula:

[0414]

[0415] The stabilization time is the total time from the moment the target detection pressure command is issued until the target pressure trajectory reaches the target detection pressure, the detection pressure difference deviation and the pressure change rate continuously meet the stabilization conditions and remain no less than the initial pressure response time constant.

[0416] The cumulative change in the target effective supply and exhaust air volume is determined by the following formula:

[0417]

[0418] Calculate the median value of the pressure overshoot, pressure stabilization time, and cumulative change of the target effective supply and exhaust air volume under each control mode;

[0419] The rate of change of the index of the control method of the present invention relative to the fixed single-branch control method is determined according to the following formula:

[0420]

[0421] In the formula: The median value of the index corresponding to the fixed single-branch control mode; This is the median value of the index corresponding to the dual-branch control method of the present invention; This represents the percentage change in the corresponding indicator.

[0422] Without changing the core control logic of the present invention, the air volume acquisition unit can adopt a direct air volume measurement device, or determine the effective supply and exhaust air volume based on the fan speed, the pressure difference before and after the fan and the pre-calibrated fan curve.

[0423] The supply and exhaust fan actuators can be bidirectional variable frequency fans, or supply fans and exhaust fans can be set separately, or a combination of constant speed fans and electric dampers can be used;

[0424] The leakage coefficient and pressure index in the dual-branch model can be refitted after obtaining new stable branch samples, but the samples of the unopened branch and the return branch after opening are saved separately, and the samples of the two branches are not mixed and fitted into a single model.

[0425] When refitting a branch model during operation, multi-objective particle swarm optimization is re-executed for the branch model that has changed; before the new pressure control parameters are generated, the original branch pressure control parameters are continued to be used, and the target pressure change rate is limited to not exceeding the upper limit of the target pressure change rate corresponding to the original control parameters;

[0426] The number of particles, maximum number of iterations, inertia coefficient and learning coefficient of the multi-objective particle swarm optimization algorithm can be adjusted according to the computing power of the industrial controller, but the particle position parameters still include the proportional control coefficient, integral control coefficient and upper limit of the target pressure change rate, and the optimization objectives still include pressure overshoot, pressure stabilization time and cumulative change of target effective supply and exhaust air volume.

[0427] The direct output of this invention is the stable establishment of the target detection pressure and the corresponding air supply and exhaust control process; the calculation of air leakage in modular buildings, the calculation of air permeability per unit area, and the evaluation of air tightness level can be performed separately after the target detection pressure is established, and are not necessary steps in the pressure control process of this invention.

[0428] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A modular building airtightness detection adaptive pressure closed-loop intelligent control system, characterized in that, include: The initial dynamic baseline establishment module is used to collect the building's internal pressure, external reference pressure, and effective supply and exhaust air volume to obtain the detection pressure difference. Through the pressure adjustment return process of the detection pressure difference, the module obtains the initial pressure response time constant, pressure repeatability boundary, pressure change rate repeatability boundary, and the set of baseline pressure and air volume differences. The bifurcation identification module is used to grade the pressure regulation return along the target detection pressure direction. Based on the pressure response time constant, pressure repeatability boundary and pressure change rate repeatability boundary, it extracts the corresponding pressure regulation branch sample and return branch sample. Based on the effective supply and exhaust air volume of the two branches and the baseline same pressure air volume difference set, it identifies the same pressure different flow return bifurcation, forming two types of branch sample sets and splice branch status. The dual-branch model establishment module is used to establish unexposed branch models and exposed return branch models based on two types of branch sample sets, and to maintain or reset the splice branch state according to the bifurcation. The swarm intelligence parameter optimization module is used to determine the corresponding pressure control parameters based on a two-branch model using a multi-objective particle swarm algorithm. The bifurcation state constraint closed-loop control module is used to select the corresponding branch model and pressure control parameters according to the splice branch state, determine the predicted leakage flow, and generate air supply and exhaust execution commands based on the predicted leakage flow and the deviation between the target detection pressure and the detection pressure difference. The pressure stability determination module is used to determine whether the target detection pressure has been stably established based on the initial pressure response time constant, pressure repeatability boundary, and pressure change rate repeatability boundary.

2. The modular building airtightness detection adaptive pressure closed-loop intelligent control system according to claim 1, characterized in that, Obtain the initial pressure response time constant, pressure repeatability boundary, pressure change rate repeatability boundary, and baseline pressure-volume difference set, including: The difference between the internal pressure of the building and the external reference pressure is defined as the detection pressure difference; Control the supply and exhaust air actuators to change the detection pressure difference along the direction of the target detection pressure until its absolute value is less than the initial pressure of the first target detection pressure, and then return under control from the initial pressure; The time required for the absolute value of the detected pressure difference to first reach 63.2% of the steady-state pressure increment during the pressure regulation process is defined as the initial pressure response time constant. Based on the median absolute deviation of the detected pressure difference and pressure change rate during the pressure holding phase in the pressure regulation process, the pressure repeatability boundary and the pressure change rate repeatability boundary are determined respectively. Based on the pressure repeatability boundary, samples in the same pressure range during the pressure regulation and return processes are paired, and the absolute values ​​of the difference between the effective supply and exhaust air volumes of the paired samples are used to form a baseline same-pressure air volume difference set.

3. The modular building airtightness detection adaptive pressure closed-loop intelligent control system according to claim 2, characterized in that, Extract samples from the voltage regulating branch and the return branch, including: Determine at least two comparison pressures from the non-final target pressures, and set a target pressure with a larger absolute value along the target detection pressure direction after each comparison pressure; When the detected differential pressure changes along the direction of the target detected pressure and enters the pressure range centered on the comparison pressure and with the pressure repeatability boundary as half the width, and the absolute value of the pressure change rate is not greater than the pressure change rate repeatability boundary, a pressure regulating sample window with a duration not less than the initial pressure response time constant is extracted, and the median value of the absolute value of the effective supply and exhaust air volume within the pressure regulating sample window is determined as the pressure regulating branch air volume. After the detected pressure difference continues to change until the absolute value is greater than the absolute value of the comparison pressure, the return is controlled. When the detected pressure difference re-enters the pressure range and meets the same conditions, the return sample window is extracted. The median value of the absolute value of the effective supply and exhaust air volume within the return sample window is determined as the return branch air volume. The difference between the return branch air volume and the pressure regulating branch air volume is determined as the return bifurcation amount of the same pressure but different flow.

4. The modular building airtightness detection adaptive pressure closed-loop intelligent control system according to claim 3, characterized in that, The identification of bifurcation of return flows with the same pressure but different flow rates includes: The sum of the absolute deviations of the median of the baseline pressure air volume difference set and three times the median after consistency correction is determined as the bifurcation decision boundary. Initialize the splice branch status to the unopened branch status; When the absolute value of the bifurcation amount of the same pressure and different flow return path is greater than the bifurcation judgment boundary, and the absolute value of the effective supply and exhaust air volume at each sampling time in the return sample window is consistent with the deviation sign of the air volume of the pressure regulating branch, it is determined that there is a bifurcation of the same pressure and different flow return path. The first determination will form a set of undiscovered branch sample samples based on at least two comparative pressures before the bifurcation of the same pressure and different flow return path; After the first determination of the bifurcation of the return flow with the same pressure and different flow, at least two stable return flow samples corresponding to the comparative pressure are used to form the sample set of the return flow branch after the opening, and the splice branch state is switched to the state of the return flow branch after the opening.

5. The modular building airtightness detection adaptive pressure closed-loop intelligent control system according to claim 4, characterized in that, Establish models for unrevealed branch lines and returned branch lines after revealing, including: When the absolute value of the pressure change rate is not greater than the repeatability boundary of the pressure change rate, the absolute value of the effective supply and exhaust air volume is determined as the observed value of the equivalent leakage flow rate. Logarithmic linear fitting was performed on the absolute values ​​of the detected pressure difference and the absolute values ​​of the effective supply and exhaust air volume in the sample sets of unopened branch lines and the sample sets of open return branch lines, respectively, to obtain the leakage coefficient and pressure index of the corresponding branch lines. A model of the unopened branch is established based on the leakage coefficient and pressure index corresponding to the unopened branch; a model of the opened return branch is established based on the leakage coefficient and pressure index corresponding to the opened return branch. The equivalent leakage flow rate determined based on the current branch model and the absolute value of the detected pressure difference will be used as the predicted leakage flow rate. Obtain the effective air volume and absolute air pressure inside the modular building. Based on the air mass conservation relationship and the ideal gas state relationship, establish a pressure dynamic model relating the rate of change of the absolute value of the detected pressure difference to the absolute value of the effective supply and exhaust air volume, the predicted leakage flow rate, the absolute air pressure, and the effective air volume inside the building.

6. The modular building airtightness detection adaptive pressure closed-loop intelligent control system according to claim 5, characterized in that, Based on the return bifurcation of the same pressure and different flow, the splice branch status is maintained or reset, including: When the splice branch state is switched to the open return branch state, the open return branch state is maintained. During subsequent pressure regulation and return processes, continue to construct same-pressure sample pairs. When the duration of the return sample window is not less than the initial pressure response time constant, and the absolute value of the same-pressure different-flow return bifurcation amount corresponding to the return sample window is not greater than the bifurcation judgment boundary, reset the splice branch state to the unopened branch state; if the reset condition is not met, continue to maintain the open return branch state.

7. The modular building airtightness detection adaptive pressure closed-loop intelligent control system according to claim 6, characterized in that, A multi-objective particle swarm optimization algorithm is used to determine the corresponding pressure control parameters based on a two-branch model, including: Multi-objective particle swarm optimization algorithm was executed for the un-uncovered branch model and the return branch model after the branch was uncovered, respectively, with the proportional control coefficient, integral control coefficient and upper limit of target pressure change rate as particle position parameters. Substitute the particle position parameters into the pressure dynamic model of the corresponding branch model, and iteratively update the particle velocity and particle position based on the particle history preferred position and the guiding particle position. The maximum difference between the absolute value of the detected pressure difference output by the pressure dynamic model and the absolute value of the target detected pressure is defined as the pressure overshoot. The time required for the detected pressure difference to first enter the pressure range centered on the target detected pressure and with the pressure repeatability boundary as half a width, and to remain there for no less than the initial pressure response time constant, is defined as the pressure stabilization time. The cumulative value of the absolute value of the change in the target effective supply and exhaust air volume obtained by the inverse solution of the pressure dynamic model during the pressure stabilization time is defined as the intensity of the change in the control quantity. With the optimization objectives of reducing pressure overshoot, stabilization time, and intensity of control quantity changes, each optimization objective is normalized, and the particle position parameter closest to the ideal target point in the normalized target space is selected to form the pressure control parameters of the unopened branch and the pressure control parameters of the return branch after opening.

8. The modular building airtightness detection adaptive pressure closed-loop intelligent control system according to claim 7, characterized in that, Based on the condition of the joint branch, select the corresponding branch model and pressure control parameters, determine the predicted leakage flow rate, and generate supply and exhaust execution commands based on the predicted leakage flow rate and the deviation between the target detection pressure and the detection pressure difference, including: Select the corresponding branch model and pressure control parameters based on the condition of the joint branch; Based on the target detection pressure, the current target trajectory pressure, and the selected upper limit of the target pressure change rate, a target pressure trajectory that changes along the target detection pressure direction is generated every control cycle. The detection pressure difference deviation is determined based on the target pressure trajectory and the detection pressure difference, and the predicted leakage flow rate is determined based on the selected branch model and the absolute value of the detection pressure difference. Based on the pressure dynamic model, predicted leakage flow, rate of change of target pressure trajectory, detection differential pressure deviation, proportional control coefficient, and integral of detection differential pressure deviation, the target effective supply and exhaust air volume is obtained by inverse solution. Based on the pre-defined correspondence between the target effective supply and exhaust air volume and the fan speed or electric damper opening, the target effective supply and exhaust air volume is converted into supply and exhaust air execution commands; When the supply and exhaust ventilation actuator reaches the execution limit, and the integral increment determined based on the detected pressure difference deviation causes the target effective supply and exhaust ventilation volume to continue to exceed the corresponding execution limit, the accumulation of integral quantity is suspended. The accumulation of integral quantity is resumed after the supply and exhaust ventilation actuator moves away from the execution limit.

9. The modular building airtightness detection adaptive pressure closed-loop intelligent control system according to claim 8, characterized in that, Determining whether the target detection pressure is stably established based on the initial pressure response time constant, pressure repeatability boundary, and pressure change rate repeatability boundary includes: The detected pressure difference is collected during the pressure closed-loop control process. The deviation of the detected pressure difference is determined based on the target pressure trajectory and the detected pressure difference, and the pressure change rate of the detected pressure difference is determined. When the target pressure trajectory reaches the target detection pressure, the absolute value of the detection pressure difference deviation is not greater than the pressure repeatability boundary, the absolute value of the pressure change rate is not greater than the pressure change rate repeatability boundary, and the duration of the above state is not less than the initial pressure response time constant, the target detection pressure is determined to be stable. Once the current target detection pressure is established stably, the system moves to the next target detection pressure, maintains or resets the joint branch status, and continues to generate air supply and exhaust execution commands based on the next target detection pressure until the control of each target detection pressure is completed.

Images