Range hood energy efficiency test method and system

Abstract

The invention discloses a range hood energy efficiency test method and system, and the method comprises the steps: selecting a typical oil smoke region set, constructing an oil stain anomaly detection model, marking abnormal points in the typical oil smoke region set based on the oil stain difference before and after oil suction, and carrying out the oil smoke test at the abnormal points. The method comprises the steps of obtaining a deviation distribution function based on a variable change condition, obtaining a time energy efficiency factor set and a space energy efficiency factor set, constructing an objective function, determining weights of factors in the time energy efficiency factor set and the space energy efficiency factor set, and constructing an energy efficiency test model based on the objective function and the weights of the factors. And obtaining test results of the range hood under different gears by using the energy efficiency test model. The method improves test comprehensiveness, improves scientificity and accuracy of product energy efficiency evaluation, has good interpretability, and can be directly applied to an energy efficiency test system of the range hood.

Description

Technical Field

[0001] The present invention relates to the field of performance testing, and in particular to a range hood energy efficiency testing method and system thereof. Background Art

[0002] As people's living standards improve, the quality of the kitchen environment has received more and more attention. As an important electrical appliance in the kitchen, the performance of the range hood directly affects the cleanliness of the kitchen air.

[0003] At present, there are many types of range hoods on the market with varying performances. Traditional range hood energy efficiency testing methods are often relatively simple and cannot fully and accurately reflect the performance of range hoods in actual use. For example, some testing methods only focus on parameters such as air volume and air pressure of range hoods, while ignoring the complexity of oil fume distribution, time and space, and the actual effects of range hoods in different usage scenarios.

[0004] In addition, with the continuous advancement of technology, new types of range hoods continue to emerge, such as smart range hoods, etc. These new products require more scientific and reasonable energy efficiency testing methods to evaluate their performance in order to meet consumers' demand for high-quality kitchen environments. Summary of the invention

[0005] The object of the present invention is to provide a range hood energy efficiency testing method and system.

[0006] To achieve the above object, the present invention is implemented according to the following technical solutions:

[0007] A first aspect of the present invention provides a range hood energy efficiency testing method, comprising:

[0008] S100 performs importance analysis on the oil fume distribution area and obtains a set of typical oil fume areas;

[0009] S200 builds an oil pollution anomaly detection model based on the isolation forest algorithm, and marks abnormal points in the oil pollution typical area set based on the difference between oil pollution before and after oil smoke absorption;

[0010] S300 obtains the change of the variable at the abnormal point by controlling a single variable, and constructs a deviation distribution function based on the change of the variable; the variable includes oil temperature and power and air volume of the range hood;

[0011] S400 uses the time domain as a calculation domain to perform a steady-state analysis on the deviation distribution function to obtain a time energy efficiency factor set, and uses the space domain as a calculation domain to perform a transient analysis on the deviation distribution function to obtain a space energy efficiency factor set;

[0012] S500 constructs an objective function, determines the weights of the factors in the time energy efficiency factor set and the space energy efficiency factor set, and constructs an energy efficiency test model based on the objective function and the weights of the factors;

[0013] S600 uses the energy efficiency test model to obtain test results of the range hood at different gears.

[0014] As a further method, the method for analyzing the importance of the oil smoke distribution area includes:

[0015] The oil fume diffusion model is constructed using fluid mechanics, and the oil fume diffusion equations are obtained, which are expressed as follows:

[0016]

[0017] Among them, ρ is the density of oil smoke, t is the operation time of the range hood, is the divergence gradient operator, is the velocity vector of the oil smoke, p is the spatial pressure, μ is the dynamic viscosity of the oil smoke, is the Laplace operator of the smoke velocity vector, is the gravitational acceleration vector, Suction for range hoods;

[0018] Obtain the distribution of oil smoke concentration, temperature and velocity at the initial moment, set boundary conditions, discretize the equation group using the finite element method, solve the discretized equation group, and obtain the distribution of oil smoke concentration, temperature and velocity at different times and locations;

[0019] The gradient values ​​of concentration, temperature and velocity distribution at different locations of the fume are calculated over time. The three gradient values ​​that are greater than the average gradient value of the diffusion process are taken as typical points, and clustering is performed to obtain a set of typical fume areas.

[0020] As a further method, the oil pollution anomaly detection model is constructed based on the isolation forest algorithm, and the method of marking abnormal points in the oil fume typical area set based on the difference in oil pollution before and after oil fume inhalation includes:

[0021] Obtain the characteristics of the oil pollution differences before and after oil fume extraction in a typical oil fume area set, including the concentration, thickness, color and texture of the oil pollution;

[0022] Select the initial partitioning features in the order of variance from large to small, and use the average value of the feature as the cutting value to divide the data space into two subspaces. Recursively repeat the partitioning in each subspace until each subspace contains only one data point, forming an isolation forest composed of multiple binary trees.

[0023] Traverse each tree in the isolation forest and calculate the anomaly score of the data point. The expression is:

[0024]

[0025] Where h is the average height of a tree in an isolation forest, Ψ is the number of trees in an isolation forest, Δξ is the offset of a data point x about the feature mean, d(x) is the traversal depth of a data point x in a tree, and w j is the weight coefficient corresponding to each layer traversal, n is the total number of data points, and γ is the Euler-Mascheroni constant;

[0026] After normalizing the anomaly scores, data points with anomaly scores in the top 10% are selected as outliers.

[0027] As a further method, the method of obtaining the change of the variable at the abnormal point by controlling the single variable and constructing the deviation distribution function based on the change of the variable includes:

[0028] Select a single variable for control, use the concentration and diffusion range of oil smoke as deviation indicators, and calculate the degree of deviation from the normal reference value for the measured value of each variable at the abnormal point;

[0029] According to the degree of deviation, the deviation distribution function is constructed through the probability density function, and the expression is:

[0030]

[0031] Among them, D(T,P,V) is the degree of deviation, T is the oil temperature, P and V are the power and air volume of the range hood respectively, |∑| is the determinant of the covariance matrix ∑, ∑ is the covariance matrix of the three variables, and θ is the mean vector.

[0032] As a further method, the method of using the time domain as a calculation domain to perform a steady-state analysis on the deviation distribution function to obtain a time energy efficiency factor set, and using the space domain as a calculation domain to perform a transient analysis on the deviation distribution function to obtain a space energy efficiency factor set includes:

[0033] The autoregressive integrated moving average model is used to model the data of the deviation distribution function in the time domain. The expression is:

[0034]

[0035] Among them, i is the current time, p is the autoregressive order, j is the lag time, φ j is the autoregressive coefficient of the jth lag time, f(D(T i-j ,P i-j ,V i-j )) is the deviation distribution function at time point ij, ∈(i) is the white noise at current time i, q is the sliding average order, θ kis the influence weight of the random error at time point ik on the current time i, ∈(ik) is the white noise at time point ik;

[0036] The time range is determined as the time interval from the start of operation of the range hood to the stable operation, and the time range is divided into a quick start period, a power adjustment period and a complete stable period based on the working characteristics of the range hood;

[0037] Calculate the steady-state mean of the model in each period and obtain the time energy efficiency factor set;

[0038] According to the main action area of ​​the range hood, the oil fume distribution area is divided into four grids: the core oil fume suction area, the near oil fume suction area, the middle distance diffusion area and the long distance edge area;

[0039] Define the spatial energy efficiency factor of the grid, the expression is:

[0040]

[0041] Where r is the index of the grid, r∈{1,2,3,4}, T r , P r and V r They are respectively represented as the oil temperature, range hood power and air volume at the corresponding grid center, f r (D(T r ,P r ,V r )) is the deviation distribution function corresponding to the grid center, ΔT r , ΔP r and ΔV r A is the change in oil temperature, range hood power and air volume at the center of the corresponding grid within 1 second. r is the volume of the corresponding grid, is the spatial gradient of the deviation distribution function;

[0042] The spatial energy efficiency factors of the four grids are combined as a spatial energy efficiency factor set.

[0043] As a further method, a method for constructing an objective function includes:

[0044] The objective function is constructed with the purpose of improving oil absorption effect and reducing power consumption. The expression is:

[0045]

[0046] Among them, V s is the actual amount of oil smoke absorbed, V t is the target value for oil fume extraction, E p is the actual oil fume purification efficiency, E pt is the target value of oil fume purification efficiency, P fis the actual fan power, P fmin and P fmax are the minimum power and maximum power of the fan respectively, E pmin and E pmax are the minimum and maximum values ​​of oil fume purification efficiency respectively.

[0047] As a further method, the method for determining the weights of the factors in the time energy efficiency factor set and the space energy efficiency factor set includes:

[0048] Conduct sensitivity analysis on the objective function to determine the impact of each factor on the objective function;

[0049] Assign initial weights of the influence ratio to the factors, substitute them into the objective function, and calculate the objective function value under the initial weights;

[0050] Use the gradient descent method to adjust the weights, calculate the gradient of the objective function for each weight, and update the weights based on the direction and magnitude of the gradient to optimize the objective function value and obtain the final weight.

[0051] As a further method, the method for constructing an energy efficiency test model based on the objective function and the weights of the factors includes:

[0052] Taking the factors as input features, an energy efficiency test model is constructed based on a neural network, including: input layer, hidden layer and output layer; the expression of the neural network is:

[0053]

[0054] Among them, y is the energy efficiency test result, g is the activation function of the output layer, w0 is the bias term, t is the number of input features, and w i is the weight of the i-th input feature to the hidden layer, x i is the i-th input feature, m is the number of neurons in the hidden layer, is the weight from the jth neuron in the hidden layer to the output layer, f is the activation function of the hidden layer, and w inj is the bias term of the jth neuron in the hidden layer, w ij is the weight of the i-th input feature of the input layer to the j-th neuron of the hidden layer;

[0055] The Markov chain Monte Carlo method is used to solve the parameters in the energy efficiency test model to maximize the posterior distribution probability of the model. The objective function is converted into the loss function of the energy efficiency test model, and the energy efficiency test model is trained to obtain the final energy efficiency test model.

[0056] As a further method, the method of using the energy efficiency test model to obtain test results of the range hood at different gears, wherein the test results are specifically: the energy efficiency test level of the range hood energy efficiency taking into account time factors and space factors.

[0057] A second aspect of the present invention provides a range hood energy efficiency testing system, comprising:

[0058] The typical area determination module is used to analyze the importance of the oil fume distribution area and obtain a set of typical oil fume areas;

[0059] An anomaly detection module, used to build an oil pollution anomaly detection model based on the isolation forest algorithm, and mark abnormal points in the oil pollution typical area set based on the difference between oil pollution before and after oil smoke absorption;

[0060] A deviation distribution module is used to obtain the change of variables at abnormal points by controlling a single variable, and construct a deviation distribution function based on the change of variables; the variables include oil temperature and the power and air volume of the range hood;

[0061] An energy efficiency index module is used to perform steady-state analysis on the deviation distribution function using the time domain as the calculation domain to obtain a time energy efficiency factor set, and to perform transient analysis on the deviation distribution function using the space domain as the calculation domain to obtain a space energy efficiency factor set;

[0062] A model building module, used to build an objective function, determine the weights of the factors in the time energy efficiency factor set and the space energy efficiency factor set, and build an energy efficiency test model based on the objective function and the weights of the factors;

[0063] The test result module is used to obtain the test results of the range hood at different gears using the energy efficiency test model.

[0064] Compared with the prior art, the embodiments of the present invention have at least the following advantages or beneficial effects:

[0065] (1) By selecting a typical oil fume area, the present invention can reduce unnecessary data processing and improve test efficiency. By selecting a typical oil fume area, the influence of some interference factors can be eliminated, so that the test results can better reflect the actual performance of the range hood.

[0066] (2) The present invention can evaluate the energy efficiency of the range hood more comprehensively by analyzing from two different dimensions, time and space. The steady-state analysis in the time domain can examine the stability and continuous performance of the range hood during a long period of operation, and understand whether its energy efficiency level is stable at different stages of use. The transient analysis in the spatial domain focuses on the range hood's ability to handle oil fumes at different spatial locations, taking into account the complexity of oil fume distribution in the kitchen space during actual use, so that the evaluation results are more in line with actual usage scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 The present invention is a flowchart of a range hood energy efficiency testing method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0068] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0069] Reference Figure 1 As shown, the present invention provides a range hood energy efficiency testing method, comprising:

[0070] S100 performs importance analysis on the oil fume distribution area and obtains a set of typical oil fume areas;

[0071] In the actual evaluation, in a standard household kitchen, professional measuring equipment was used to collect the concentration, temperature and velocity distribution data of the oil smoke at the initial moment. The oil smoke concentration near the stove was measured to be 85mg / m 3 The concentration in the center of the kitchen is 60 mg / m 3 . The temperature was measured using an infrared thermometer. The temperature above the stove was 130°C and 90°C a little further away. The velocity distribution was determined using a small wind speed meter. The wind speed near the window was 0.4m / s and above the stove was 0.25m / s. Reasonable boundary conditions were set based on the actual situation of the kitchen. The kitchen wall was set to have a fume velocity of 0 and a temperature that was the same as the indoor temperature of 25°C. The finite element method was then used to discretize and solve the fume diffusion equations to obtain specific values ​​at different times and locations. After the range hood had been running for 2 minutes, the fume concentration at the center of the kitchen became 40mg / m 3 , temperature is 100℃, speed is 0.15m / s;

[0072] In the actual evaluation, by calculating the gradient values ​​of concentration, temperature and velocity distribution at each position in time, the average gradient value of the diffusion process is obtained as the concentration gradient 8mg / m 3 / min, temperature gradient 8℃ / min, velocity gradient 0.08m / s / min, and concentration gradient at one location is 12mg / m 3 / min, temperature gradient is 10℃ / min, speed gradient is 0.12m / s / min, this position is determined as a typical point. After cluster analysis, the area with a radius of 0.6 meters around the stove is finally determined as a typical area set of oil smoke;

[0073] S200 builds an oil pollution anomaly detection model based on the isolation forest algorithm, and marks abnormal points in the oil pollution typical area set based on the difference between oil pollution before and after oil smoke absorption;

[0074] It should be explained that the Isolation Forest Algorithm can effectively identify abnormal points that are different from the normal oil pollution data pattern. By comparing the oil pollution status of a typical oil smoke area set before and after oil extraction, it can accurately determine which areas have poor oil pollution removal effects. Its role is to help us quickly locate the parts of the range hood that may have problems so that targeted improvements and optimizations can be made. It also provides key abnormal data references for subsequent more accurate evaluation of the energy efficiency of the range hood;

[0075] It should be understood that when constructing an isolation forest, the algorithm mainly relies on the path length of the tree, rather than the actual value of the data, to calculate the anomaly score. This is because outliers are usually more different from normal data points and are more easily isolated;

[0076] In the actual evaluation, the oil pollution characteristics of each location in the typical oil fume area set were measured before and after the range hood was working. Before the oil fume was sucked, the oil pollution concentration in the typical area was 1.8 mg / cm 2 , with a thickness of 0.08 mm, dark brown in color, and a relatively rough texture. After the range hood has been working for a period of time, the oil concentration in this area becomes 1 mg / cm 2 , the thickness is 0.04mm, the color becomes lighter to light brown, and the texture is relatively smoother; the feature variance is analyzed in descending order, and it is found that the variance of the color feature is the largest after calculation, so the color is selected as the initial partition feature, and the average color value is used as the cutting value to divide the data space into two subspaces. Assuming that the average color value is a specific grayscale value of 115, color values ​​greater than 115 are divided into one subspace, and those less than 115 are divided into another subspace, and then the division is continued recursively until each subspace contains only one data point, and finally an isolated forest composed of multiple binary trees is formed; then traverse each tree in the isolated forest, a total of 80 trees, calculate the anomaly score of the data point, and after normalizing the anomaly scores of all data points, select the data points with anomaly scores in the top 10% as anomalies;

[0077] S300 obtains the change of the variable at the abnormal point by controlling a single variable, and constructs a deviation distribution function based on the change of the variable; the variable includes oil temperature and power and air volume of the range hood;

[0078] It needs to be explained that "obtaining the change of variables at abnormal points by controlling a single variable" means changing only one variable of the oil temperature and the power and air volume of the range hood each time, and observing the impact of the change of this variable at the marked abnormal point, while "constructing a deviation distribution function based on the change of variables" is to quantify the degree of deviation of this change from the normal state, so as to conduct subsequent energy efficiency analysis;

[0079] In the actual evaluation, the oil temperature was selected as a single variable for control. The normal reference oil temperature was 95°C, and the oil temperature measured at the abnormal point was 105°C. At the same time, the oil fume concentration measured at this location was 35 mg / m 3 , the diffusion range is 0.25 meters in radius. By calculating the degree of deviation from the normal reference value, the probability density function is used to construct the deviation distribution function. Assuming that the covariance matrix determinant is 0.04, the covariance matrix is The mean vector is [100, 1400, 18] (corresponding to oil temperature, range hood power, and air volume, respectively);

[0080] S400 uses the time domain as a calculation domain to perform a steady-state analysis on the deviation distribution function to obtain a time energy efficiency factor set, and uses the space domain as a calculation domain to perform a transient analysis on the deviation distribution function to obtain a space energy efficiency factor set;

[0081] It should be explained that by analyzing the energy efficiency index values ​​at different time stages, the energy efficiency change trend of the range hood in the time dimension can be obtained. This change trend can help determine the performance of the range hood in different gears, such as the start-up speed and stable operation time; by analyzing the spatial energy efficiency index values ​​in each grid area, the energy efficiency change trend of the range hood in the spatial dimension can be obtained;

[0082] In the actual evaluation, the autoregressive integral moving average model is used to model the data of the deviation distribution function in the time domain. The autoregressive order is 2, the moving average order is 1, the lag time is 1 minute and 2 minutes respectively, the autoregressive coefficient is [0.55, 0.35], the white noise of the current time is 0.08, the weight of the random error at the time point on the current time is 0.35, and the white noise at the time point is 0.15; the time interval from the start of the range hood to the stable operation is determined to be 4 minutes, and the time range is divided into a rapid start-up period (0-1 minute), a power adjustment period (1-2.5 minutes) and a completely stable period (2.5-4 minutes) based on the working characteristics of the range hood. By calculating the steady-state mean of the model in each period, the time energy efficiency factor set is obtained, with the mean of the rapid start-up period being 0.75, the mean of the power adjustment period being 1.15, and the mean of the completely stable period being 1.4;

[0083] In the actual evaluation, the oil fume distribution area is divided into four grids according to the main working area of ​​the range hood: the core oil fume absorption area, the near oil fume absorption area, the medium-distance diffusion area and the long-distance edge area. The oil temperature at the center of the core oil fume absorption area grid is 108°C, the range hood power is 1700W, and the air volume is 19m 3 / min, the deviation distribution function value is 0.65, the oil temperature changes by 1.8℃ within 1s, the range hood power changes by 45W, and the air volume changes by 0.8m 3 / min, volume is 0.9m 3 , the spatial gradient of the deviation distribution function is 0.45;

[0084] S500 constructs an objective function, determines the weights of the factors in the time energy efficiency factor set and the space energy efficiency factor set, and constructs an energy efficiency test model based on the objective function and the weights of the factors;

[0085] In the actual evaluation, the objective function is constructed with the purpose of improving the oil absorption effect and reducing power consumption. The factors are oil temperature, range hood power, air volume, 3 factors in the time energy efficiency factor set and 4 factors in the space energy efficiency factor set. The gradient descent method is used to adjust the weights. After multiple iterations, the gradient of the objective function to each weight is calculated. The weight is updated based on the direction and size of the gradient to make the objective function value optimal and obtain the final weight value of the factor. These factors are used as input features to build an energy efficiency test model based on a neural network. The Markov chain Monte Carlo method is used to solve the parameters in the energy efficiency test model and train the energy efficiency test model.

[0086] S600 uses the energy efficiency test model to obtain test results of the range hood at different gears.

[0087] In the actual evaluation, the energy efficiency test model was used to obtain the test results of the range hood at different gears. The test results were specifically the energy efficiency test level of the range hood energy efficiency that combined time factors and space factors. It was found that at the high gear, the energy efficiency test level was 82%, at the medium gear it was 72%, and at the low gear it was 62%.

[0088] The second aspect of the present invention further provides a range hood energy efficiency testing system, comprising:

[0089] The typical area determination module is used to analyze the importance of the oil fume distribution area and obtain a set of typical oil fume areas;

[0090] An anomaly detection module, used to build an oil pollution anomaly detection model based on the isolation forest algorithm, and mark abnormal points in the oil pollution typical area set based on the difference between oil pollution before and after oil smoke absorption;

[0091] A deviation distribution module is used to obtain the change of variables at abnormal points by controlling a single variable, and construct a deviation distribution function based on the change of variables; the variables include oil temperature and the power and air volume of the range hood;

[0092] An energy efficiency index module is used to perform steady-state analysis on the deviation distribution function using the time domain as the calculation domain to obtain a time energy efficiency factor set, and to perform transient analysis on the deviation distribution function using the space domain as the calculation domain to obtain a space energy efficiency factor set;

[0093] A model building module, used to build an objective function, determine the weights of the factors in the time energy efficiency factor set and the space energy efficiency factor set, and build an energy efficiency test model based on the objective function and the weights of the factors;

[0094] The test result module is used to obtain the test results of the range hood at different gears using the energy efficiency test model.

[0095] In this embodiment, the method for analyzing the importance of oil smoke distribution areas includes:

[0096] The oil fume diffusion model is constructed using fluid mechanics, and the oil fume diffusion equations are obtained, which are expressed as follows:

[0097]

[0098] Among them, ρ is the density of oil smoke, t is the operation time of the range hood, is the divergence gradient operator, is the velocity vector of the oil smoke, p is the spatial pressure, μ is the dynamic viscosity of the oil smoke, is the Laplace operator of the smoke velocity vector, is the gravitational acceleration vector, Suction for range hoods;

[0099] Obtain the distribution of oil smoke concentration, temperature and velocity at the initial moment, set boundary conditions, discretize the equation group using the finite element method, solve the discretized equation group, and obtain the distribution of oil smoke concentration, temperature and velocity at different times and locations;

[0100] The gradient values ​​of concentration, temperature and velocity distribution at different locations of the fume are calculated over time. The three gradient values ​​that are greater than the average gradient value of the diffusion process are taken as typical points, and clustering is performed to obtain a set of typical fume areas.

[0101] In this embodiment, the oil pollution anomaly detection model is constructed based on the isolation forest algorithm, and the method for marking abnormal points in the oil fume typical area set based on the difference in oil pollution before and after oil fume inhalation includes:

[0102] Obtain the characteristics of the oil pollution differences before and after oil fume extraction in a typical oil fume area set, including the concentration, thickness, color and texture of the oil pollution;

[0103] Select the initial partitioning features in the order of variance from large to small, and use the average value of the feature as the cutting value to divide the data space into two subspaces. Recursively repeat the partitioning in each subspace until each subspace contains only one data point, forming an isolation forest composed of multiple binary trees.

[0104] Traverse each tree in the isolation forest and calculate the anomaly score of the data point. The expression is:

[0105]

[0106] Where h is the average height of a tree in an isolation forest, Ψ is the number of trees in an isolation forest, Δξ is the offset of a data point x about the feature mean, d(x) is the traversal depth of a data point x in a tree, and w j is the weight coefficient corresponding to each layer traversal, n is the total number of data points, and γ is the Euler-Mascheroni constant;

[0107] After normalizing the anomaly scores, data points with anomaly scores in the top 10% are selected as outliers.

[0108] In this embodiment, the method of obtaining the change of the variable at the abnormal point by controlling a single variable and constructing a deviation distribution function based on the change of the variable includes:

[0109] Select a single variable for control, use the concentration and diffusion range of oil smoke as deviation indicators, and calculate the degree of deviation from the normal reference value for the measured value of each variable at the abnormal point;

[0110] According to the degree of deviation, the deviation distribution function is constructed through the probability density function, and the expression is:

[0111]

[0112] Among them, D(T,P,V) is the degree of deviation, T is the oil temperature, P and V are the power and air volume of the range hood respectively, |∑| is the determinant of the covariance matrix ∑, ∑ is the covariance matrix of the three variables, and θ is the mean vector.

[0113] In this embodiment, the method of performing steady-state analysis on the deviation distribution function using the time domain as the calculation domain to obtain the time energy efficiency factor set, and performing transient analysis on the deviation distribution function using the space domain as the calculation domain to obtain the space energy efficiency factor set includes:

[0114] The autoregressive integrated moving average model is used to model the data of the deviation distribution function in the time domain. The expression is:

[0115]

[0116] Among them, i is the current time, p is the autoregressive order, j is the lag time, φ j is the autoregressive coefficient of the jth lag time, f(D(T i-j ,P i-j ,V i-j )) is the deviation distribution function at time point ij, ∈(i) is the white noise at current time i, q is the sliding average order, θ k is the influence weight of the random error at time point ik on the current time i, ∈(ik) is the white noise at time point ik;

[0117] The time range is determined as the time interval from the start of operation of the range hood to the stable operation, and the time range is divided into a quick start period, a power adjustment period and a complete stable period based on the working characteristics of the range hood;

[0118] Calculate the steady-state mean of the model in each period and obtain the time energy efficiency factor set;

[0119] According to the main action area of ​​the range hood, the oil fume distribution area is divided into four grids: the core oil fume suction area, the near oil fume suction area, the middle distance diffusion area and the long distance edge area;

[0120] Define the spatial energy efficiency factor of the grid, the expression is:

[0121]

[0122] Where r is the index of the grid, r∈{1,2,3,4}, T r , P r and V r They are respectively represented as the oil temperature, range hood power and air volume at the corresponding grid center, f r (D(T r ,P r ,V r )) is the deviation distribution function corresponding to the grid center, ΔT r , ΔP r and ΔV r A is the change in oil temperature, range hood power and air volume at the center of the corresponding grid within 1 second. r is the volume of the corresponding grid, is the spatial gradient of the deviation distribution function;

[0123] The spatial energy efficiency factors of the four grids are combined as a spatial energy efficiency factor set.

[0124] In this embodiment, the method for constructing the objective function includes:

[0125] The objective function is constructed with the purpose of improving oil absorption effect and reducing power consumption. The expression is:

[0126]

[0127] Among them, V s is the actual amount of oil smoke absorbed, V t is the target value for oil fume extraction, E p is the actual oil fume purification efficiency, E pt is the target value of oil fume purification efficiency, P f is the actual fan power, P fmin and P fmax are the minimum power and maximum power of the fan respectively, E pmin and E pmax are the minimum and maximum values ​​of oil fume purification efficiency respectively.

[0128] In this embodiment, the method for determining the weights of the factors in the time energy efficiency factor set and the space energy efficiency factor set includes:

[0129] Conduct sensitivity analysis on the objective function to determine the impact of each factor on the objective function;

[0130] Assign initial weights of the influence ratio to the factors, substitute them into the objective function, and calculate the objective function value under the initial weights;

[0131] Use the gradient descent method to adjust the weights, calculate the gradient of the objective function for each weight, and update the weights based on the direction and magnitude of the gradient to optimize the objective function value and obtain the final weight.

[0132] In this embodiment, the method for constructing an energy efficiency test model based on an objective function and factor weights includes:

[0133] Taking the factors as input features, an energy efficiency test model is constructed based on a neural network, including: input layer, hidden layer and output layer; the expression of the neural network is:

[0134]

[0135] Among them, y is the energy efficiency test result, g is the activation function of the output layer, w0 is the bias term, t is the number of input features, and w i is the weight of the i-th input feature to the hidden layer, x i is the i-th input feature, m is the number of neurons in the hidden layer, is the weight from the jth neuron in the hidden layer to the output layer, f is the activation function of the hidden layer, and w inj is the bias term of the jth neuron in the hidden layer, w ijis the weight of the i-th input feature of the input layer to the j-th neuron of the hidden layer;

[0136] The Markov chain Monte Carlo method is used to solve the parameters in the energy efficiency test model to maximize the posterior distribution probability of the model. The objective function is converted into the loss function of the energy efficiency test model, and the energy efficiency test model is trained to obtain the final energy efficiency test model.

[0137] In this embodiment, the method of using the energy efficiency test model to obtain test results of the range hood at different gears, wherein the test results are specifically: the energy efficiency test level of the range hood energy efficiency taking into account time factors and space factors.

[0138] The above contents are merely examples and explanations of the structure of the present invention. The technicians in this technical field may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the structure of the invention or exceed the scope defined by the claims, they should all fall within the protection scope of the present invention.

Claims

1. A range hood energy efficiency testing method, characterized in that: The following steps are involved: Carry out importance analysis on the oil fume distribution area and obtain a set of typical oil fume areas; An oil pollution anomaly detection model is constructed based on the isolation forest algorithm, and abnormal points are marked in the oil pollution typical area set based on the difference in oil pollution before and after oil fume absorption; The change of the variable is obtained at the abnormal point by controlling a single variable, and a deviation distribution function is constructed based on the change of the variable; the variable includes oil temperature and power and air volume of the range hood; Taking the time domain as the calculation domain, the deviation distribution function is subjected to steady-state analysis to obtain the time energy efficiency factor set; taking the space domain as the calculation domain, the deviation distribution function is subjected to transient analysis to obtain the space energy efficiency factor set; Constructing an objective function, determining weights of factors in the time energy efficiency factor set and the space energy efficiency factor set, and constructing an energy efficiency test model based on the objective function and the weights of the factors; The energy efficiency test model is used to obtain test results of the range hood at different gears.

2. A range hood energy efficiency testing method according to claim 1, characterized in that: The method for analyzing the importance of oil smoke distribution areas comprises: The oil fume diffusion model is constructed using fluid mechanics, and the oil fume diffusion equations are obtained, which are expressed as follows: Among them, ρ is the density of oil smoke, t is the operation time of the range hood, is the divergence gradient operator, is the velocity vector of the oil smoke, p is the spatial pressure, μ is the dynamic viscosity of the oil smoke, is the Laplace operator of the smoke velocity vector, is the gravitational acceleration vector, Suction for range hoods; Obtain the distribution of oil smoke concentration, temperature and velocity at the initial moment, set boundary conditions, discretize the equation group using the finite element method, solve the discretized equation group, and obtain the distribution of oil smoke concentration, temperature and velocity at different times and locations; The gradient values ​​of concentration, temperature and velocity distribution at different locations of the fume over time are calculated, and the three gradient values ​​that are greater than the average gradient value of the diffusion process are taken as typical points, and clustering is performed to obtain a set of typical fume areas.

3. A range hood energy efficiency testing method according to claim 1, characterized in that: The method of constructing an oil pollution anomaly detection model based on the isolation forest algorithm and marking abnormal points in the oil pollution typical area set based on the difference between oil pollution before and after oil smoke inhalation includes: Obtain the characteristics of the oil pollution differences before and after oil fume extraction in a typical oil fume area set, including the concentration, thickness, color and texture of the oil pollution; Select the initial partitioning features in the order of variance from large to small, and use the average value of the feature as the cutting value to divide the data space into two subspaces. Recursively repeat the partitioning in each subspace until each subspace contains only one data point, forming an isolation forest composed of multiple binary trees. Traverse each tree in the isolation forest and calculate the anomaly score of the data point. The expression is: Where h is the average height of a tree in an isolation forest, Ψ is the number of trees in an isolation forest, Δξ is the offset of a data point x about the feature mean, d(x) is the traversal depth of a data point x in a tree, and w j is the weight coefficient corresponding to each layer traversal, n is the total number of data points, and γ is the Euler-Mascheroni constant; After normalizing the anomaly scores, data points with anomaly scores in the top 10% are selected as outliers.

4. A range hood energy efficiency testing method according to claim 1, characterized in that: The method of obtaining the change of the variable at the abnormal point by controlling a single variable and constructing a deviation distribution function based on the change of the variable includes: Select a single variable for control, use the concentration and diffusion range of oil smoke as deviation indicators, and calculate the degree of deviation from the normal reference value for the measured value of each variable at the abnormal point; According to the degree of deviation, the deviation distribution function is constructed through the probability density function, and the expression is: Among them, D(T,P,V) is the degree of deviation, T is the oil temperature, P and V are the power and air volume of the range hood respectively, |∑| is the determinant of the covariance matrix ∑, ∑ is the covariance matrix of the three variables, and θ is the mean vector.

5. A range hood energy efficiency testing method according to claim 1, characterized in that: The method of using the time domain as a calculation domain to perform steady-state analysis on the deviation distribution function to obtain a time energy efficiency factor set, and using the space domain as a calculation domain to perform transient analysis on the deviation distribution function to obtain a space energy efficiency factor set, comprises: The autoregressive integrated moving average model is used to model the data of the deviation distribution function in the time domain. The expression is: Among them, i is the current time, p is the autoregressive order, j is the lag time, φ j is the autoregressive coefficient of the jth lag time, f(D(T i-j ,P i-j ,V i-j )) is the deviation distribution function at time point ij, ∈(i) is the white noise at current time i, q is the sliding average order, θ k is the influence weight of the random error at time point ik on the current time i, ∈(ik) is the white noise at time point ik; The time range is determined as the time interval from the start of operation of the range hood to the stable operation, and the time range is divided into a quick start period, a power adjustment period and a complete stable period based on the working characteristics of the range hood; Calculate the steady-state mean of the model in each period and obtain the time energy efficiency factor set; According to the main action area of ​​the range hood, the oil fume distribution area is divided into four grids: the core oil fume suction area, the near oil fume suction area, the middle distance diffusion area and the long distance edge area; Define the spatial energy efficiency factor of the grid, the expression is: Where r is the index of the grid, r∈{1,2,3,4}, T r , P r and V r They are respectively represented as the oil temperature, range hood power and air volume at the corresponding grid center, f r (D(T r ,P r ,V r )) is the deviation distribution function corresponding to the grid center, ΔT r , ΔP r and ΔV r A is the change in oil temperature, range hood power and air volume at the center of the corresponding grid within 1 second. r is the volume of the corresponding grid, is the spatial gradient of the deviation distribution function; The spatial energy efficiency factors of the four grids are combined as a spatial energy efficiency factor set.

6. A range hood energy efficiency testing method according to claim 1, characterized in that: The method for constructing the objective function comprises: The objective function is constructed with the purpose of improving oil absorption effect and reducing power consumption. The expression is: Among them, V s is the actual amount of oil smoke absorbed, V t is the target value for oil fume extraction, E p is the actual oil fume purification efficiency, E pt is the target value of oil fume purification efficiency, P f is the actual fan power, P fmin and P fmax are the minimum power and maximum power of the fan respectively, E pmin and E pmax are the minimum and maximum values ​​of oil fume purification efficiency respectively.

7. A range hood energy efficiency testing method according to claim 1, characterized in that: The method for determining the weights of the factors in the time energy efficiency factor set and the space energy efficiency factor set comprises: Conduct sensitivity analysis on the objective function to determine the impact of each factor on the objective function; Assign initial weights of the influence ratio to the factors, substitute them into the objective function, and calculate the objective function value under the initial weights; Use the gradient descent method to adjust the weights, calculate the gradient of the objective function for each weight, and update the weights based on the direction and magnitude of the gradient to optimize the objective function value and obtain the final weight.

8. The range hood energy efficiency testing method according to claim 1, characterized in that: The method for constructing an energy efficiency test model based on an objective function and factor weights includes: Taking the factors as input features, an energy efficiency test model is constructed based on a neural network, including: input layer, hidden layer and output layer; the expression of the neural network is: Among them, y is the energy efficiency test result, g is the activation function of the output layer, w0 is the bias term, t is the number of input features, and w i is the weight of the i-th input feature to the hidden layer, x i is the i-th input feature, m is the number of neurons in the hidden layer, is the weight from the jth neuron in the hidden layer to the output layer, f is the activation function of the hidden layer, and w inj is the bias term of the jth neuron in the hidden layer, w ij is the weight of the i-th input feature of the input layer to the j-th neuron of the hidden layer; The Markov chain Monte Carlo method is used to solve the parameters in the energy efficiency test model to maximize the posterior distribution probability of the model. The objective function is converted into the loss function of the energy efficiency test model, and the energy efficiency test model is trained to obtain the final energy efficiency test model.

9. A range hood energy efficiency testing method according to claim 1, characterized in that: The method of using the energy efficiency test model to obtain test results of the range hood at different gears, wherein the test results are specifically: the energy efficiency test level of the range hood energy efficiency that integrates time factors and space factors.

10. A range hood energy efficiency testing system, used to execute a range hood energy efficiency testing method according to any one of claims 1 to 9, characterized in that: The system comprises: The typical area determination module is used to analyze the importance of the oil fume distribution area and obtain a set of typical oil fume areas; An anomaly detection module, used to build an oil pollution anomaly detection model based on the isolation forest algorithm, and mark abnormal points in the oil pollution typical area set based on the difference between oil pollution before and after oil smoke absorption; A deviation distribution module is used to obtain the change of variables at abnormal points by controlling a single variable, and construct a deviation distribution function based on the change of variables; the variables include oil temperature and the power and air volume of the range hood; An energy efficiency index module is used to perform steady-state analysis on the deviation distribution function using the time domain as the calculation domain to obtain a time energy efficiency factor set, and to perform transient analysis on the deviation distribution function using the space domain as the calculation domain to obtain a space energy efficiency factor set; A model building module, used to build an objective function, determine the weights of the factors in the time energy efficiency factor set and the space energy efficiency factor set, and build an energy efficiency test model based on the objective function and the weights of the factors; The test result module is used to obtain the test results of the range hood at different gears using the energy efficiency test model.

Images