Airtightness detection method and device adaptive to new energy vehicle electric cabinet

Abstract

The invention discloses an air tightness detection method and device adaptive to an electric cabinet of a new energy vehicle, and relates to the technical field of air tightness detection. Comprising the following steps: fixing an electric cabinet in a self-adaptive clamp, and driving a flexible sealing gasket to be attached to the surface of the electric cabinet through a corner cylinder to form a closed detection cavity; and gas with set pressure of 10-50 kpa is filled into the closed green lifting device, pressure data in the cavity are collected in real time through a high-precision gas pressure sensor, and the collection frequency is larger than or equal to 100 Hz. A fuzzy PID control algorithm is adopted, the inflation rate and the pressure maintaining time are dynamically adjusted according to the real-time leakage rate, accurate pressure control in the detection process is guaranteed, meanwhile, historical leakage data are analyzed through a pre-trained LSTM neural network model, the future leakage trend is predicted, a detection report containing a three-dimensional leakage thermodynamic diagram is generated, and the detection accuracy is improved. The leakage data is analyzed by using the LSTM neural network model, an accurate air tightness monitoring result and leakage trend prediction are provided, and accurate evaluation is provided for the air tightness of the electric cabinet.

Description

Technical Field

[0001] The present invention relates to the technical field of airtightness detection, and in particular to an airtightness detection method and device adapted to an electric control box of a new energy vehicle. Background Art

[0002] The electric control box of new energy vehicles (also called electric control system, electronic control unit or electronic control box) is one of the most important components in new energy vehicles. It is mainly responsible for controlling and managing the work of electric motors, power batteries, charging systems and other related systems.

[0003] In the prior art, air tightness detection mainly adopts two methods: static test and dynamic test. Static test usually makes judgment by observing the rate of pressure drop after inflation and pressurization. However, this method needs to wait for a long time for the air pressure to stabilize during the test, and is easily affected by environmental changes, resulting in unstable test results. Although dynamic test can improve the detection rate to a certain extent, it still faces large errors, especially for small leaks of precision components that are difficult to accurately capture.

[0004] Therefore, the present invention further improves the accuracy and efficiency of detection by combining fuzzy PID control and LSTM neural network model to meet the current new energy vehicles' increasing demand for high sealing and rapid detection of electric control boxes. Summary of the invention

[0005] The purpose of the present invention is to provide an airtightness detection method and device adapted for an electric control box of a new energy vehicle, so as to solve the problems raised in the above-mentioned background technology.

[0006] To achieve the above purpose, the present invention provides the following technical solution: a method and device for detecting airtightness of an electric control box adapted to a new energy vehicle, comprising the following steps:

[0007] Fix the electric control box in the adaptive fixture, and drive the flexible sealing pad to fit on the surface of the electric control box through the angle cylinder to form a closed detection cavity;

[0008] Fill the sealed green grape with a set pressure of 10-50kpa, and collect the pressure data in the cavity in real time through a high-precision air pressure sensor, with a collection frequency of ≥100Hz;

[0009] Based on the pressure decay curve fitting equation , combined with the cavity solvent V and time t, calculate the leakage rate ,in is the standard atmospheric pressure, and is compared with the preset threshold Q threshold to determine air tightness;

[0010] Fuzzy PID control algorithm is used to dynamically adjust the inflation rate and pressure holding time according to the real-time leakage rate. The control parameters include the proportional coefficient , Integration time , Differential time ;

[0011] Analyze leakage data based on the pre-trained LSTM neural network model, predict leakage trends and generate detection reports including 3D leakage heat maps.

[0012] As a specific solution of the technical solution of this application, the adaptive fixture adjusts the fixture spacing through the XYZ three-axis slide rail driven by the servo motor, and the size range of the adapted electric control box is 80% to 120% of the standard specification;

[0013] The flexible sealing gasket is made of silicone rubber and has a corrugated structure on the surface. The contact pressure is controlled within the range of 0.5-2Mpa through closed-loop feedback.

[0014] As a specific solution of the technical solution of the present application, the pressure decay curve fitting adopts the least square method, and the fitting error threshold is set to ±0.05%FS;

[0015] The leakage rate threshold Q threshold is dynamically adjusted according to the IP protection level of the electric control box: IP67 level corresponds to Q threshold = 0.5Pa / min, and IP68 level corresponds to Q threshold = 0.2Pa / min.

[0016] As a specific solution of the technical solution of this application, the fuzzy PID control algorithm includes:

[0017] The input variables are the real-time leakage rate error e(t) and its change rate ec(t), and the output is the inflation valve opening adjustment Δu(t);

[0018] The fuzzy rule base contains 25 control rules. If e(t) is positive and large and ec(t) is negative and small, then Δu(t) is negative and medium. The defuzzification adopts the centroid method, and the output resolution is ≤0.1%.

[0019] As a specific solution of the technical solution of this application, the construction of the LSTM neural network model includes:

[0020] Input layer: time series pressure data P(t), cavity volume V, ambient temperature T;

[0021] Hidden layer: 3 layers of LSTM units, with 128, 64, and 32 neurons in each layer respectively;

[0022] Output layer: leakage probability distribution map and predicted leakage rate Q pred , the training loss function adopts mean square error, and the number of training iterations is ≥ 1000 times.

[0023] As a specific solution of the technical solution of the present application, it also includes the cavity volume self-calibration, which measures the outer dimensions of the electric control box by a laser rangefinder before detection, and calculates the actual cavity volume V in combination with the CAD model. 实际 , volume calibration error ≤ 0.1%;

[0024] Environmental compensation, based on the temperature sensor and humidity sensor data, linear compensation is performed on the leakage rate calculation results. The compensation formula is: , where a=0.003 / ℃ is the temperature coefficient and β=0.001 / %RH is the humidity coefficient.

[0025] As a specific solution of the technical solution of this application, the inflation process is divided into two stages: the first stage: inflation at a rate of 10 kpa / s to 90% of the target pressure;

[0026] The second stage: Inflate to the target pressure at a rate of 2 kPa / s, and control the overshoot to ≤0.5%.

[0027] An airtightness detection device adapted for an electric control box of a new energy vehicle, the airtightness detection method for the electric control box using any one of claims 1 to 7.

[0028] Compared with the prior art, the present invention has the following beneficial effects:

[0029] The airtightness detection method and device adapted to the electric control box of new energy vehicles adopts a fuzzy PID control algorithm to dynamically adjust the inflation rate and the pressure holding time according to the real-time leakage rate to ensure accurate pressure control during the detection process. At the same time, the pre-trained LSTM neural network model is used to analyze historical leakage data, predict future leakage trends, and generate a detection report containing a three-dimensional leakage heat map. By using the LSTM neural network model to analyze leakage data, accurate airtightness monitoring results and leakage trend predictions are provided, thereby providing an accurate evaluation of the airtightness of the electric control box. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a schematic diagram of the flow of the method for detecting air tightness of an electric control box of the present invention. DETAILED DESCRIPTION

[0031] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0032] In the description of the present invention, it is to be understood that the terms “center”, “longitudinal”, “lateral”, “length”, “width”, “thickness”, “up”, “down”, “front”, “back”, “left”, “right”, “vertical”, “horizontal”, “top”, “bottom”, “inside”, “outside”, “clockwise”, “counterclockwise”, “axial”, “radial”, “circumferential”, etc., indicating orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the present invention.

[0033] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of the present invention, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0034] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements, unless otherwise clearly defined. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0035] In the present invention, unless otherwise clearly specified and limited, a first feature being "above" or "below" a second feature may mean that the first and second features are in direct contact, or the first and second features are in indirect contact through an intermediate medium. Moreover, a first feature being "above", "above" or "above" a second feature may mean that the first feature is directly above or obliquely above the second feature, or simply means that the first feature is higher in level than the second feature. A first feature being "below", "above" or "below" a second feature may mean that the first feature is directly above or obliquely above the second feature, or simply means that the first feature is lower in level than the second feature.

[0036] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0037] like Figure 1 As shown, the present invention provides a technical solution: a method and device for detecting airtightness of an electric control box adapted to a new energy vehicle, comprising the following steps:

[0038] Fix the electric control box in the adaptive fixture, and drive the flexible sealing pad to fit on the surface of the electric control box through the angle cylinder to form a closed detection cavity;

[0039] Fill the sealed green grape with a set pressure of 10-50kpa, and collect the pressure data in the cavity in real time through a high-precision air pressure sensor, with a collection frequency of ≥100Hz;

[0040] Based on the pressure decay curve fitting equation , combined with the cavity solvent V and time t, calculate the leakage rate ,in is the standard atmospheric pressure, and is compared with the preset threshold Q threshold to determine air tightness;

[0041] Fuzzy PID control algorithm is used to dynamically adjust the inflation rate and pressure holding time according to the real-time leakage rate. The control parameters include the proportional coefficient , Integration time , Differential time ;

[0042] Analyze leakage data based on the pre-trained LSTM neural network model, predict leakage trends and generate detection reports including 3D leakage heat maps. It should be clear that in the embodiment of the present application, the electric control box is first accurately fixed in the adaptive clamp, and the adaptive clamp can automatically adjust the clamping force and position according to the shape and size of the electric control box. The adaptive clamp is a prior art and will not be described in detail in the present application. In the present application, the flexible sealing gasket is driven by the angle cylinder to fit the surface of the electric control box. The angle cylinder can provide a stable driving force to make the flexible sealing gasket evenly fit various parts of the surface of the electric control box to form a closed detection cavity. In the present application, the sealing gasket is made of oil-resistant, wear-resistant and high-temperature resistant rubber material. In the present application, the gas is usually selected from dry, clean air or inert gases such as nitrogen. The inflation pressure is determined according to the actual working pressure and design requirements of the electric control box. This pressure range can ensure the accuracy of the detection while avoiding damage to the internal components of the electric control box due to excessive pressure. Inflation is achieved by a high-precision gas booster pump, and the inflation rate must be strictly controlled during the inflation process to ensure that the pressure in the cavity increases evenly and steadily. It should also be clear that in the present application, a high-precision air pressure sensor is used to collect pressure data in the cavity in real time, and the collection frequency is ≥100Hz. The high-precision air pressure sensor can accurately sense the pressure changes in the cavity and convert them into electrical signals. The high collection frequency can ensure that enough pressure data points are obtained to accurately draw the pressure decay curve. For example, 100 or more pressure data can be collected within 1 second, so the pressure changes in a short period of time can be recorded in detail. In the present application, the calculated leakage rate Q is compared with the preset threshold value Qthreshold. If Q ≤ Qthreshold, the air tightness of the electric control box is determined to be qualified. If Q > Qthreshold, the air tightness is determined to be unqualified. The preset threshold value Qthreshold is determined based on factors such as the design requirements of the electric control box and the actual working environment, and represents the maximum allowable leakage rate that the electric control box can meet normal use requirements.As can be seen from the previous text, the inflation rate and pressure holding time adjustment based on the fuzzy PID control algorithm, the fuzzy PID control algorithm is a control method that combines fuzzy logic with traditional PID (proportional-integral-differential) control, which can dynamically adjust the inflation rate and pressure holding time according to the actual leakage rate to improve the efficiency and accuracy of the detection process. The proportional coefficient kp determines the response speed of the controller to the current error, the integral time Ti is used to eliminate the steady-state error of the system, and the differential time Td can predict the future trend of the system and make adjustments in advance. Then, according to the real-time leakage rate, the fuzzy rules are used to adjust the proportional coefficient, integral time and differential time. For example, when the real-time leakage rate is large, the proportional coefficient Kp can be appropriately increased to speed up the inflation rate so that the pressure in the cavity reaches the preset value as soon as possible. At the same time, the integral time Ti is appropriately reduced and the differential time Td is increased to improve the stability and response speed of the system. Through the fuzzy PID control algorithm, the inflation rate and pressure holding time can be finely controlled to ensure effective airtightness detection under different leakage conditions. It should be clear that in the embodiment of the present application, the leakage data analysis and report generation based on the LSTM neural network model first need to collect a large amount of leakage data, including leakage rates, pressure changes and other data of different electric control boxes under different working conditions, for training the LSTM neural network model. During the training process, the leakage data is input into the LSTM network, and the network weights and bias parameters are adjusted to make the prediction results output by the network as close as possible to the actual leakage situation. Then the leakage data collected in real time is input into the pre-trained LSTM neural retention model. The model can analyze the leakage data and predict the leakage trend, for example, predict whether the leakage rate will continue to increase, decrease or remain stable during the subsequent detection process. According to the prediction results, a test report containing a three-dimensional leakage heat map is generated. The three-dimensional leakage heat map can intuitively display the leakage of various parts of the electric control box, where the horizontal and vertical coordinates can represent the plane position of the electric control box, the third dimension (height direction) can represent the size of the leakage rate, and different colors are used to distinguish different leakage degrees. The test report can also include the air tightness determination result of the electric control box, the specific location and severity of the leakage, and other information, providing a basis for maintenance and improvement.

[0043] The adaptive fixture adjusts the fixture spacing through the XYZ three-axis slide rail driven by the servo motor, and the size range of the adapted electric control box is 80% to 120% of the standard specification;

[0044] The flexible sealing gasket is made of silicone rubber and has a corrugated structure on the surface. The contact pressure is controlled within the range of 0.5-2Mpa through closed-loop feedback.

[0045] The least square method was used for fitting the pressure decay curve, and the fitting error threshold was set at ±0.05%FS;

[0046] The leakage rate threshold Q threshold is dynamically adjusted according to the IP protection level of the electric control box: IP67 level corresponds to Q threshold = 0.5Pa / min, and IP68 level corresponds to Q threshold = 0.2Pa / min. It should be clear that in the embodiment of the present application, the pressure decay curve fitting adopts the least squares method to find the best function match for the data. In this detection method, by collecting the data points of the pressure change in the cavity over time (P 1 , t 1 )、(P 2 , t2), .... (P n , t n ), construct the mathematical model of the pressure decay curve, as shown in Formula 1: , where P(t) represents the pressure in the cavity at time t, P 0 represents the initial pressure, K represents the pressure attenuation coefficient, and the collected data are fitted by the least squares method to solve P 0 The values ​​of and k are set so that the error score and k between the fitting curve and the actual data points are minimized. The fitting error threshold is set to ±0.05%FS, where FS stands for full scale, i.e., the maximum pressure range that the detection system can measure. For example, if the pressure measurement range of the detection system is 0-100kpa, then the fitting error threshold is ±0.05×100kPa = ±0.05kPa. During the fitting process, if the error between the calculated pressure value and the actual measured value exceeds this threshold, data collection and fitting are required again. In this application, the leakage rate threshold Q threshold is dynamically adjusted according to the IP protection level of the electric control box. The IP protection level refers to the protection level of the housing of the electrical equipment against the intrusion of foreign objects. Different IP protection levels have different requirements for the sealing performance of the equipment. For the IP67 level, the corresponding Q threshold = 0.5Pa / min. The IP67 level requires that the equipment can be immersed in water for a certain time (usually 30 minutes) and a certain depth without entering the water. Under this protection level, the operating leakage rate of the electric control box is extremely large. Therefore, the Q threshold is set to 0.5Pa / min. During the detection process, if the calculated leakage rate exceeds 0.5Pa / min, the sealing performance of the inventory electric control box does not meet the IP67 requirements. For the IP68 level, the corresponding Q threshold = 0.2Pa / min, the IP68 level has higher protection requirements than the IP67 level, and the equipment can remain sealed for a longer time and in more complex situations. Therefore, its operating leakage rate is lower, and the Q threshold is set to 0.2Pa / min. During the detection, only when the leakage rate does not exceed 0.2Pa / min, the electric control box is judged to meet the IP68 sealing performance requirements. It can be seen from the pressure decay curve equation formula 1 that the pressure decay coefficient k is related to the leakage rate Q. By differentiating formula 1, the pressure change rate can be obtained: , according to the definition of leakage rate, the leakage rate Q can be expressed as: In the detection process, assuming that the pressure change is small in a short time range, it can be approximated ,therefore , it shows that the pressure attenuation coefficient k is proportional to the leakage rate Q. After obtaining k through fitting, the leakage rate Q can be further calculated.

[0047] The fuzzy PID control algorithm includes:

[0048] The input variables are the real-time leakage rate error e(t) and its change rate ec(t), and the output is the inflation valve opening adjustment Δu(t);

[0049] The fuzzy rule base contains 25 control rules. If e(t) is positive and large and ec(t) is negative and small, then Δu(t) is negative and medium. The defuzzification adopts the centroid method, and the output resolution is ≤0.1%. It should be clear that in the embodiment of the present application, the real-time leakage rate error e(t) is the difference between the expected leakage rate (usually 0, indicating no leakage) and the actual real-time leakage rate Q(t), that is, e(t) = Q expected - Q(t). For example, if the expected leakage rate is 0pa / min and the actual real-time leakage rate is 0.3kp / min, then e(t) = -0.3kp / min. The leakage rate error change rate ec(t) is the derivative of the error e(t) with respect to time t, that is, ec(t) = de(t) / dt, which reflects the speed of the error change over time. For example, at two consecutive sampling times t 1 and t 2 (interval is Δt = 1s), the errors are e(t 1 ) = 0.2Pa / min and e(t 2 ) = 0.3Pa / min, then ec(t) ≈ (0.3 - 0.2) / 1 = 0.1Pa / min². For the input variables e(t) and ec(t), and the output variable Δu(t), fuzzy sets are defined respectively, such as {negative large (NB), negative small (NS), zero (ZO), positive small (PS), positive large (PB)}. Each fuzzy set corresponds to a membership function, which is used to determine the degree to which the input or output variable belongs to the fuzzy set. For example, for e(t), its membership function can be defined as a triangular function or a bell-shaped function, etc., and the parameters of the function are determined based on actual experience and experiments, so that the error value can be reasonably mapped to the fuzzy set. In this application, the defuzzification adopts the centroid method, also known as the area center method. It regards the membership function of the fuzzy output set as a curve, calculates the centroid position of the area enclosed by the curve and the horizontal axis, and thus obtains an accurate output value Δu(t). The specific calculation formula is [Formula 1]: , where u is the possible value of the output variable inflation valve opening adjustment, and μΔu(u) is the membership function of Δu(u).

[0050] The construction of the LSTM neural network model includes:

[0051] Input layer: time series pressure data P(t), cavity volume V, ambient temperature T;

[0052] Hidden layer: 3 layers of LSTM units, with 128, 64, and 32 neurons in each layer respectively;

[0053] Output layer: leakage probability distribution map and predicted leakage rate Q pred , the training loss function adopts mean square error, and the number of training iterations is ≥1000 times. It should be clear that in the embodiment of the present application, the pressure data in the cavity collected in real time during the detection process takes the time series as input. For example, assuming that the acquisition frequency is 100Hz and the detection time is 10 seconds, the length of the input pressure data sequence is 10*100=1000 data points. During the detection process, the ambient temperature can be monitored in real time by a temperature sensor, or in a relatively stable detection environment, an average temperature value can be taken as input, for example, T=25°C. In the present application, the training loss function adopts mean square error to measure the difference between the predicted output (including the leakage probability distribution diagram and the predicted leakage rate) and the actual target value. The calculation formula of the mean square error is Formula 2: , where n is the number of samples and y true is the actual target value (including the actual leakage probability distribution diagram and the actual leakage rate), y pred It is the model prediction output value. For example, assuming there is a set of training data, the true leakage rate y_true_Q = 0.5Pa / min, and the model predicted leakage rate y_pred_Q = 0.45Pa / min, then the mean square error of the sample is (0.5 - 0.45)² = 0.0025. For the leakage probability distribution map, assuming that the true leakage probability of a certain area is 0.8 and the predicted probability is 0.7, then the mean square error of the area is (0.8 - 0.7)² = 0.01. In this application, the number of iterations refers to the number of repetitions of the forward propagation and back propagation process on the entire training data set. A sufficient number of iterations can enable the model to fully learn the rules in the data, reduce the value of the loss function, and improve the accuracy and generalization ability of the model. In the actual training process, you can decide whether to stop training in advance based on the changes in the loss function. For example, when the loss function changes very little or tends to be stable in multiple consecutive iterations, you can stop training to save computing resources and prevent overfitting.

[0054] Cavity volume self-calibration: Before testing, the outer dimensions of the electric control box are measured by a laser rangefinder, and the actual cavity volume V is calculated in combination with the CAD model. 实际 , volume calibration error ≤ 0.1%;

[0055] Environmental compensation, based on the temperature sensor and humidity sensor data, linear compensation is performed on the leakage rate calculation results. The compensation formula is: , where a=0.003 / ℃ is the temperature coefficient, and β=0.001 / %RH is the humidity coefficient. In the embodiment of the present application, the standard error of the volume is required to be ≤0.1%, then the error between the calculated cavity volume and the actual volume shall not exceed 0.1%. For example, if the actual volume is 0.1m³, the calibrated volume error shall not exceed 0.1×0.1m³ = 0.0001. During the detection process, the ambient temperature T and humidity H data are collected in real time through the temperature sensor and the humidity sensor. The measuring range of the temperature sensor is usually -50℃-100℃, and the accuracy can reach ±0.1℃; the measuring range of the humidity sensor is 0%~100%RH, and the accuracy can reach ±1%RH. For example, the collected ambient temperature T=25℃, the humidity H=50%RH, and the leakage rate calculation result is linearly compensated according to the temperature coefficient a and the humidity coefficient β. The compensation formula is Formula 1: Qcomp = Qmeasured + a×(T - T 0 )+β×(H - H 0 ), where Qcomp is the compensated leakage rate, Qmeasured is the actual measured leakage rate, a = 0.003 / ℃ is the temperature coefficient, β =0.001 / %RH is the humidity coefficient, T 0 and H 0 are the reference temperature and humidity (usually the initial ambient temperature and humidity during the test). For example, assuming the actual measured leakage rate Qmeasured = 0.3Pa / min, the reference temperature T 0 =20℃, reference humidity H 0 = 40%RH, the current ambient temperature T = 25℃, and the humidity H = 50%RH, then the compensated leakage rate is: Qcomp = 0.3 + 0.003×(25 - 20) + 0.001×(50 - 40) = 0.3 + 0.015 + 0.01 =0.325Pa / min. The above cavity volume self-calibration and environmental compensation measures can improve the accuracy and reliability of the airtightness detection of the electric control box of new energy vehicles, ensuring that the test results are not affected by environmental factors and volume measurement errors.

[0056] The inflation process is divided into two stages: the first stage: inflation at a rate of 10kpa / s to 90% of the target pressure;

[0057] The second stage: Inflate to the target pressure at a rate of 2 kPa / s, and control the overshoot to ≤0.5%.

[0058] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is limited by the attached embodiments and their equivalents.

Claims

1. A method for detecting air tightness of an electric control box adapted to a new energy vehicle, characterized in that: The following steps are involved: Fix the electric control box in the adaptive fixture, and drive the flexible sealing pad to fit on the surface of the electric control box through the angle cylinder to form a closed detection cavity; Fill the sealed green grape with a set pressure of 10-50kpa, and collect the pressure data in the cavity in real time through a high-precision air pressure sensor, with a collection frequency of ≥100Hz; Based on the pressure decay curve fitting equation , combined with the cavity solvent V and time t, calculate the leakage rate ,in is the standard atmospheric pressure, and is compared with the preset threshold Q threshold to determine air tightness; Fuzzy PID control algorithm is used to dynamically adjust the inflation rate and pressure holding time according to the real-time leakage rate. The control parameters include the proportional coefficient , Integration time , Differential time ; Analyze leakage data based on the pre-trained LSTM neural network model, predict leakage trends and generate detection reports including 3D leakage heat maps.

2. According to claim 1, a method for detecting air tightness of an electric control box adapted to a new energy vehicle is characterized in that: The adaptive fixture adjusts the fixture spacing through the XYZ three-axis slide rail driven by the servo motor, and the size range of the adapted electric control box is 80% to 120% of the standard specification; The flexible sealing gasket is made of silicone rubber and has a corrugated structure on the surface. The contact pressure is controlled within the range of 0.5-2Mpa through closed-loop feedback.

3. According to claim 1, a method for detecting air tightness of an electric control box adapted to a new energy vehicle is characterized in that: The pressure decay curve is fitted using the least square method, and the fitting error threshold is set to ±0.05%FS; The leakage rate threshold Q threshold is dynamically adjusted according to the IP protection level of the electric control box: IP67 level corresponds to Q threshold = 0.5Pa / min, and IP68 level corresponds to Q threshold = 0.2Pa / min.

4. According to claim 1, a method for detecting air tightness of an electric control box adapted to a new energy vehicle is characterized in that: The fuzzy PID control algorithm includes: The input variables are the real-time leakage rate error e(t) and its change rate ec(t), and the output is the inflation valve opening adjustment Δu(t); The fuzzy rule base contains 25 control rules. If e(t) is positive and large and ec(t) is negative and small, then Δu(t) is negative and medium. The defuzzification adopts the centroid method, and the output resolution is ≤0.1%.

5. The airtightness detection method for an electric control box adapted to a new energy vehicle according to claim 1 is characterized in that: The construction of the LSTM neural network model includes: Input layer: time series pressure data P(t), cavity volume V, ambient temperature T; Hidden layer: 3 layers of LSTM units, with 128, 64, and 32 neurons in each layer respectively; Output layer: leakage probability distribution map and predicted leakage rate Q pred , the training loss function adopts mean square error, and the number of training iterations is ≥ 1000 times.

6. The method for detecting air tightness of an electric control box adapted to a new energy vehicle according to claim 1 is characterized in that: It also includes the cavity volume self-calibration, which uses a laser rangefinder to measure the size of the electric control box before detection, and calculates the actual cavity volume V in combination with the CAD model. 实际 , volume calibration error ≤ 0.1%; Environmental compensation, based on the temperature sensor and humidity sensor data, linear compensation is performed on the leakage rate calculation results. The compensation formula is: , where a=0.003 / ℃ is the temperature coefficient and β=0.001 / %RH is the humidity coefficient.

7. The method for detecting air tightness of an electric control box adapted to a new energy vehicle according to claim 1 is characterized in that: The inflation process is divided into two stages: the first stage: inflation at a rate of 10 kPa / s to 90% of the target pressure; The second stage: Inflate to the target pressure at a rate of 2 kPa / s, and control the overshoot to ≤0.5%.

8. An airtightness detection device adapted to the electric control box of a new energy vehicle, characterized in that: The airtightness detection method of the electric control box uses any one of claims 1 to 7.

Images