Intelligent canteen management system and method

Abstract

The invention discloses an intelligent canteen management system and method, and relates to the technical field of canteen management. Dinner plate waste data is acquired, the food waste amount is calculated, a dinner plate waste data set is established, a dinner plate classification model is constructed by using a twin neural network algorithm, and after the dinner plate data is screened, a waste analysis table is constructed; the method comprises the following steps: establishing a waste analysis index WI analysis model based on a waste analysis table, determining a key analysis window, analyzing related data of the key window, drawing a curve, judging the trend of the curve by using Hilbert-Huang transform, judging the change trend of the waste index by using a cumulative first-order difference ratio algorithm if the trend changes are consistent, and adjusting the food supply quantity of the dinner plate. The system comprises a data acquisition module, a data reconstruction module, a window analysis module, a trend analysis module and an intelligent regulation and control module, multiple factors are integrated, an effective means is provided for intelligent canteen management, and reduction of food waste is facilitated.

Description

Technical Field

[0001] The present invention relates to the technical field of canteen management, and in particular to a smart canteen management system and method. Background Art

[0002] In the current canteen management field, the problem of food waste is becoming increasingly serious. Traditional canteen management methods lack effective monitoring, analysis and regulation methods in dealing with food waste.

[0003] The existing technology simply counts the amount of food left over, but fails to deeply analyze the complex factors behind the waste data. For example, it fails to distinguish whether the food wasted on the plates comes from the canteen supply window, resulting in poor data accuracy and difficulty in accurately locating the root cause of the problem. When evaluating the waste situation at canteen supply windows, existing technologies mostly only consider the single factor of average waste amount, ignoring the comprehensive impact of key factors such as waste differences in different time periods, the discrete degree of plate waste, and fluctuations in the number of plates on the overall waste situation. It is unable to fully reflect the actual waste level of each window, and it is difficult to effectively guide canteen managers to formulate targeted waste reduction strategies.

[0004] To this end, the present invention provides a smart canteen management system and method. Summary of the invention

[0005] The object of the present invention is to provide a smart canteen management system and method to solve at least one of the above-mentioned prior art problems.

[0006] In a first aspect, the present invention provides a smart canteen management method, comprising: Step 1: Obtain the meal tray data from the cafeteria supply window in the cafeteria and establish a meal tray waste data group; Step 2: Based on the plate waste data set, build a plate classification model, determine the waste data of the plates that can be analyzed, remove invalid data in the plate waste data set, and build a waste analysis table X ij ; Step 3: Based on the waste analysis table X ij , establish a waste analysis index WI analysis model, sort and analyze the waste analysis index WI of each canteen supply window, and determine the key analysis window; Step 4: Based on the key analysis window, numerical analysis is performed on the number of plates in the monitoring period to obtain the plate value for the period, and curve analysis is performed on the waste analysis index and plate value for the period of multiple monitoring periods to determine whether the change trend of the curve is consistent; Step 5. If the waste analysis index and time plate values ​​of multiple monitoring cycles have the same changing trend, determine whether the change in the waste analysis index shows an increasing trend. If it is increasing, adjust the food supply of each plate in the key analysis window to achieve intelligent management of canteen waste.

[0007] In a second aspect, the present invention provides a smart canteen management system, comprising: Data acquisition module: obtains the plate data from the canteen supply window in the canteen and establishes a plate waste data group; Data reconstruction module: Based on the plate waste data group, a plate classification model is constructed to determine and analyze the waste data of the plates that can be used, remove invalid data in the plate waste data group, and construct a waste analysis table; Window analysis module: Based on waste analysis table X ij , establish a waste analysis index WI analysis model, sort and analyze the waste analysis index WI of each canteen supply window, and determine the key analysis window; Trend analysis module: Based on the key analysis window, numerical analysis is performed on the number of plates in the monitoring period to obtain the plate value for the period. Curve analysis is performed on the waste analysis index and plate value for the period of multiple monitoring periods to determine whether the changing trends of the curves are consistent. Intelligent control module: If the waste analysis index and time period plate values ​​of multiple monitoring cycles have the same curve change trend, determine whether the waste analysis index change shows an increasing trend. If it is increasing, adjust the food supply of each plate in the key analysis window to achieve intelligent management of canteen waste.

[0008] Beneficial effects of the present invention: 1. Through the plate data collection mechanism, such as embedding RFID tags and multi-device collaborative information acquisition, combined with the twin neural network algorithm to build a plate classification model, distinguish the sources of food waste on the plates, ensure the high accuracy and reliability of the data, and establish a comprehensive waste analysis index WI model that comprehensively considers multiple factors and determines the key analysis window, so that canteen managers can get to the core of the waste problem and lay the foundation for subsequent precise policy implementation.

[0009] 2. Use Hilbert-Huang transform and other advanced technologies to conduct curve analysis and trend judgment on monitoring data, capture the changing dynamics of waste analysis index and time period plate value, and activate early warning and intelligent control mechanism when the waste index is found to have an increasing trend, dynamically adjust the food supply of key window plates, which is conducive to dealing with waste problems. It can also be flexibly adapted according to the actual operation status of the canteen to realize the management of smart canteen. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0011] Figure 1 It is a flow chart of a smart canteen management method of the present invention; Figure 2 This is a module intention of a smart canteen management system provided by Embodiment 3 of the present invention; Figure 3 It is a structural diagram of a computer device provided in Embodiment 4 of the present invention. DETAILED DESCRIPTION

[0012] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme 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, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0013] Embodiment 1 Figure 1 This is a flowchart of a smart canteen management method provided in the first embodiment of the present invention. The embodiment of the present invention can be applied to the situation of food waste management in a smart canteen. The smart canteen management method can be executed by a smart canteen management system, which can be implemented by software and / or hardware, and can be configured in a computer device. Optionally, the computer device can be an electronic device, which can be a notebook, a desktop computer, a smart tablet, etc., and the embodiment of the present invention does not limit this.

[0014] like Figure 1 As shown, a smart canteen management method provided by an embodiment of the present invention includes the following steps: Step 1: Obtain the meal tray data from the cafeteria supply window in the cafeteria and establish a meal tray waste data group; In some embodiments, an RFID tag is embedded in each meal plate, and the canteen supply window and the meal plate recycling window obtain the meal plate data through a weighing device and an image acquisition device, and write the meal plate data into the smart canteen database; The plate data includes: the weight of each plate, the plate number, the canteen supply window number corresponding to the plate number, and the link address of the plate image; It should be noted that the plate recycling window removes debris other than food from the plates to be recycled to improve the accuracy of the plate weight, the image acquisition device matches the plate image with the plate number one by one, and stores the plate image in the server, and the server stores the link address of the server of the plate image in the smart canteen database; Obtain the same plate weight at the canteen supply window and the plate recycling window from the smart canteen database, perform difference processing on the plate weights at the canteen supply window and the plate recycling window to obtain the amount of food waste; Obtain the amount of food waste for each plate at a single canteen supply window during the monitoring period, and use the amount of food waste for each plate as a data element to establish a plate waste data group; Step 2: Based on the plate waste data set, build a plate classification model, determine the waste data of the plates that can be analyzed, remove invalid data in the plate waste data set, and build a waste analysis table X ij ; Obtain images of canteen trays at the canteen supply window and tray recycling window from the smart canteen database and server, and use the twin neural network algorithm to build a tray classification model; Based on the plate classification model, the images of plates at the canteen supply window and plate recycling window are analyzed to determine whether the wasted food on the plates comes from the canteen supply window; It should be noted that the twin neural network algorithm is a special neural network architecture that contains two identical sub-networks, which are used to extract feature vectors of the images of the dining trays at the cafeteria supply window and the dining tray recycling window respectively; The similarity of images is determined by comparing the Euclidean distance or cosine similarity of feature vectors. During the training phase, it is necessary to provide the supply and recovery images of the same plate as positive samples of the twin neural network algorithm, and the images of different plates as negative samples for twin neural network training. Specifically, the plate classification model is constructed as follows: Mark the images of the dining trays at the canteen supply window and the dining tray recycling window as supply images and recycling images respectively; S1. Data preparation and preprocessing: Obtain images of meal trays at the canteen supply window and meal tray recycling window during the monitoring period. Use the supply and recycling images of the same meal tray as positive samples of the twin neural network algorithm, and the images of different meal trays as negative samples. Standardize the meal tray images and normalize the pixel values ​​of the images to the range of [0,1]. S2, model architecture construction stage, PyTorch is selected as the deep learning framework, two identical sub-networks with convolutional neural network as the core are constructed, the convolution layer, pooling layer, and fully connected layer of the plate classification model are constructed, and the feature vector of the plate image is output at the end of the sub-network, and the Euclidean distance and cosine similarity of the feature vector are calculated to output the similarity of the feature vector; Among them, the convolution layer uses a small convolution kernel of 3x3 and a maximum pooling layer of 2x2; It should be noted that the plate image contains rich details such as the plate outline, food shape, and tableware style. Small convolution kernels can capture these local features. Selecting the 2x2 maximum pooling layer can halve the size of the feature map while retaining significant features, reduce the amount of calculation, and speed up the model operation. Fully connected layer: The fully connected layer plays the role of feature dimensionality reduction and integration, mapping it to a low-dimensional space, which is convenient for the subsequent direct calculation of the similarity of the output feature vectors of the two sub-networks; S3, select the contrast loss function, use the standardized plate images and the corresponding similarity data as the data set, divide the training set and the test set in a ratio of 7:3, and start the plate classification model training; S4, evaluating the results output by the plate classification model, and outputting the F1 value of the plate classification model. If the F1 value is higher than 0.85, the task plate classification model training is qualified, otherwise continue to train the model; Through the plate classification model, the same plate supply image and recycling image are output to the plate classification model to determine whether the wasted food on the plate comes from the same supply window of the cafeteria; If the wasted plates come from the same canteen supply window, mark the plates as usable plates for analysis, otherwise mark them as unusable plates for analysis; The amount of food waste that cannot be analyzed on plates is removed from the plate waste data group to construct a waste analysis data group; Obtain the waste analysis data set of all canteen supply windows in the canteen and construct the waste analysis table X ij , i is the number of the plate waste data group, and the value range of i is [1,n j ], j represents the number of the canteen supply window, and the value range of j is [1,m]; Each row of the waste analysis table contains the amount of food waste in a waste analysis data group of a canteen supply window; It should be noted that, due to the different lengths of the waste analysis data sets of different canteen supply windows, the upper limit of the range of i is n j , some canteen supply windows will fill in the missing data in the waste analysis table with 0.

[0015] Embodiment 2 like Figure 1 As shown, a smart canteen management method also includes the following steps: Step 3: Based on the waste analysis table X ij , establish a waste analysis index WI analysis model, sort and analyze the waste analysis index WI of each canteen supply window, and determine the key analysis window; Based on the waste analysis table X ij , establish the waste analysis index WI (Waste Index) analysis model; Specifically, the average waste amount AW of the waste analysis data group in each canteen supply window is calculated j ; By formula: Get the average waste amount AW of the waste analysis data group in each canteen supply window j； By formula: Get the coefficient of variation CW of each canteen supply window j ; The sub-index WI1 based on the average waste amount and dispersion degree is established by the formula: , get the sub-index WI of the average waste and dispersion degree of each canteen supply window 1j ; Waste Analysis Table X ij According to the different food supply periods in the cafeteria, multiple analysis periods k are divided, k=1, 2, 3, 4, 5; For example, k=1, 2, 3, 4, 5 correspond to the breakfast period, lunch period, afternoon tea period, dinner period, and supper period of the cafeteria respectively; Obtain the amount of wasted plates in the same analysis period of the canteen supply window and sum them up to obtain the amount of wasted plates in each analysis period W jk , j is the number of the canteen supply window, and k is the analysis period; By formula: Get the proportion of plate waste in each analysis period P jk ; The sub-index WI2 based on the analysis period is established by the formula: , where w k is the proportionality factor for each analysis period, w k They are: 0.25, 0.43, 0.12, 0.56, 0.28; It should be noted that the sub-index WI of the analysis period 2j Considering the waste in different time periods and the importance of each time period, the sub-index WI of the time period is analyzed. 2j It can highlight the impact of waste during key meal periods on the overall window waste. For example, if the proportion coefficient of the dinner period is high, and the waste ratio of a canteen supply window during the dinner period is also high, then the canteen supply window will be more wasteful during the WI 2j The scores in the sub-index will increase accordingly, helping to identify cafeteria supply windows with high waste during important meal periods; Sub-index WI based on meal period 2j and the average waste and dispersion sub-index WI 1j , through weighted summation calculation, we get the window waste index WI 3j ; Establish a fluctuating waste index WI for each canteen supply window 4j ; By formula: Get the fluctuation waste index WI 4j , where X ij max、X ij min are waste analysis table X ij The maximum and minimum values ​​of plate waste at the same cafeteria supply window; Get the window waste index WI of each canteen supply window 3j , the fluctuation waste index WI of each canteen supply window 4j , the window waste index WI 3j , Fluctuation waste index WI 4j Perform product processing to obtain the waste analysis index WI of each canteen supply window; Obtain the waste analysis index WI of each canteen supply window, arrange the waste analysis index WI of each canteen supply window in descending order, and determine the canteen supply window with the largest waste analysis index WI value as the key analysis window; It should be noted that the waste analysis index WI is obtained by numerically calculating the window waste index and the fluctuation waste index. The waste analysis index WI combines the average waste amount, the degree of dispersion, the waste situation in different time periods and the fluctuation range of the waste amount. The waste analysis index WI can comprehensively and comprehensively measure the waste situation of each supply window, thereby determining the key analysis window with the most serious waste, and providing canteen managers with a unified and comprehensive evaluation standard so that they can take effective management measures to reduce food waste.

[0016] Step 4: Based on the key analysis window, numerical analysis is performed on the number of plates in the monitoring period to obtain the plate value for the period, a plate change model is constructed, and curve analysis is performed on the waste analysis index and plate value for the period of multiple monitoring periods to determine whether the change trend of the curve is consistent; Based on the key analysis window, the waste analysis index WI in multiple monitoring cycles is obtained, and the monitoring cycle-waste analysis index WI sequence W(t) is constructed; Obtain the number of plates in each analysis period of the key analysis window within each monitoring cycle, sum and average the number of plates in each analysis period within the monitoring cycle, and obtain the plate value for the period; Obtain meal plate values ​​for multiple monitoring periods and construct a monitoring period-period meal plate value sequence S(t); It should be noted that the time period plate average can reflect the sales volume changes in the key analysis window; In the two-dimensional rectangular coordinate system, the change curves of the monitoring cycle-waste analysis index WI and the monitoring cycle-time period plate average are drawn; The Hilbert-Huang transform (HTT) was performed on the monitoring cycle-waste analysis index WI sequence W(t) and the monitoring cycle-time plate value sequence S(t) to construct a plate change analysis model; Specifically, the plate change analysis model is constructed as follows: SS1. Find extreme points and construct envelopes: For the monitoring cycle-waste analysis index WI sequence W(t), find all local extreme points and construct upper and lower envelopes by cubic spline interpolation method; For the monitoring cycle-waste analysis index WI sequence W(t), find all the local extreme points of the sequence, where the local extreme points include: local maximum points and local minimum points; After finding all the extreme points, the upper and lower envelopes are constructed by cubic spline interpolation. Cubic spline interpolation is a mathematical method that can generate two smooth curves based on given extreme points, which serve as the upper envelope and the lower envelope respectively. It should be noted that the upper and lower envelopes are used to "wrap" the original time series and can reflect the approximate range and trend of the monitoring cycle-waste analysis index WI series W(t) and the monitoring cycle-time plate value series S(t); SS2, calculate the average value e(t) of the upper and lower envelopes, and the difference Cz(t) between the monitoring cycle-waste analysis index WI sequence W(t) and the average value e(t) of the upper and lower envelopes of the average value, and obtain the difference sequence; Calculate the average values ​​of the upper and lower envelopes, as well as the difference Cz(t) of the monitoring cycle-waste analysis index WI sequence W(t), Cz(t) = |W(t)-e(t)|, and construct the monitoring cycle-difference sequence based on the difference Cz(t); SS3, determine whether the monitoring period-difference sequence meets the judgment condition of the intrinsic mode function IMF. If it meets the judgment condition of the monitoring period-waste analysis index WI sequence W(t), extract the remaining sequence and repeat the decomposition of the remaining sequence; Judgment condition 1: Monitoring period - waste analysis index WI sequence W(t), the number of extreme value points and the number of zero crossing points are equal or differ by at most 1; Judgment condition 2: At the monitoring time within any monitoring period, the average value of the upper and lower envelopes constructed by the local maximum point and the local minimum point is 0; If the monitoring cycle-waste analysis index WI sequence W(t) meets the IMF judgment condition, then the monitoring cycle-difference sequence is an IMF, that is, an intrinsic mode function. If not, the monitoring cycle-difference sequence is used as a new time series, and SS1 and SS2 are repeated until the first intrinsic mode function IMF is obtained, marked as IMF1(t); It should be noted that IMF in Hilbert-Huang Transform (HHT) refers to Intrinsic Mode Function (IMF), which is a function that satisfies certain conditions; By formula: Get the remaining sequence r1(t); It should be noted that the residual series r1(t) contains other information in the original time series besides the oscillation pattern of the represented IMF1(t); Take the remaining sequence r1(t) as the new original sequence, repeat SS2 and SS3 until the remaining sequence r s (t) satisfies the cut-off condition, and we get IMF2(t), IMF3(t), ..., IMF s (t), and construct the remaining sequence , r2(t), ..., r s (t), the total number of decompositions of s; Among them, the cut-off conditions include: the remaining sequence r s (t) Become a monotonic function or the number of extreme points is less than 3. Monotonic functions include: monotonic increasing function and monotonic decreasing function; When the remaining sequence reaches the cutoff condition, the monitoring cycle-waste analysis index WI sequence W(t) is reconstructed and marked as , , where h is the number of the IMF, and the value range of h is [1, s], where r s (t) is the trend component of the monitoring cycle-waste analysis index WI series W(t) or the residual part that cannot be decomposed any further; SS4, yes Perform Hilbert transform (HT) and calculate each The instantaneous phase θ h (t); right Performing Hilbert transform, we get , for The conjugate function of ; calculate and The instantaneous phase θ h (t), ; Based on the monitoring cycle-waste analysis index WI sequence W(t) instantaneous phase θ h (t) acquisition method, get the monitoring cycle-time plate value sequence S (t) instantaneous phase ; SS5. Calculate the monitoring cycle - instantaneous phase θ of the waste analysis index WI sequence W(t) h (t), monitoring cycle-period plate value sequence S(t) instantaneous phase Phase difference , determine whether the change curves of the monitoring cycle-waste analysis index WI and the monitoring cycle-time period plate average are synchronous changes; By formula: , obtain the monitoring cycle-waste analysis index WI sequence W(t) instantaneous phase θ h (t), the phase difference of the instantaneous phase of the monitoring period-time plate value sequence S (t) ; If the phase difference of the monitoring cycle exist Within the range, the monitoring period is marked as the synchronous change period, and the phase difference of the synchronous change period is marked as ; During the acquisition monitoring period, the phase difference is not The monitoring period within the range is marked as an asynchronous change period, and the asynchronous change period is marked as ; Phase difference based on synchronous change cycle , the phase difference of the asynchronous change cycle , calculate the synchronous trend value Tb; By formula: Obtain the synchronous trend value Tb, where Nt represents the total number of synchronous change cycles, Yt represents the total number of asynchronous change cycles, t represents the monitoring period, and d1 and d2 are 0.378 and 0.564 respectively; Compare the synchronous trend value Tb with the synchronous trend threshold to determine whether the waste analysis index and the time period meal plate value of multiple monitoring cycles have the same change trend; If the synchronous trend value Tb is higher than the synchronous trend threshold, it indicates that the waste analysis index and the plate value of the time period in multiple monitoring cycles have the same change trend of the curves, otherwise, they are inconsistent; It should be noted that if the two curves have the same changing trends, for example, as the monitoring period progresses, the waste analysis index WI and the average value of plates in the time period increase at the same time, it means that in the key analysis window, the increase in plate sales is accompanied by a simultaneous deterioration in food waste. For example, in order to satisfy more customers, the speed of serving meals is increased, but the amount of meals served is not well controlled, resulting in a simultaneous increase in waste and plate sales. On the contrary, if the two curves drop at the same time, it may be that the window has taken effective management measures to reduce food waste, such as reducing waste caused by oversupply, and adjusting the menu structure so that customers can choose food more reasonably; When the two curves move in opposite directions, for example, the waste analysis index WI increases while the average number of plates in a period decreases, this may mean that although the number of customers at the focus analysis window, that is, the sales volume of plates, is decreasing, the food waste situation is getting worse. This may be because the window did not adjust the dish supply strategy in time when the number of customers decreased, resulting in an increase in the waste of remaining food, or because the service quality declined, which made the remaining customers less satisfied with the food, resulting in more waste; On the contrary, if the waste analysis indicates that WI decreases while the average number of plates in a period increases, this may be because the window has taken some measures to reduce waste while increasing sales, so that food waste is effectively controlled.

[0017] Step 5: If the waste analysis indexes and plate values ​​of multiple monitoring cycles have the same change trend, determine whether the waste analysis index changes in an increasing trend. If so, adjust the food supply of each plate in the key analysis window to achieve intelligent management of canteen waste; If the curve change trends of the waste analysis index WI and the plate value of the time period are consistent in multiple monitoring periods, determine the overall change trend of the waste analysis index or the plate value curve of the time period, and determine whether the overall trend of the curve tends to increase or decrease; The first-order difference of the monitoring cycle-waste analysis index WI series W(t) is calculated by the cumulative first-order difference ratio algorithm. , , Indicates the number of the monitoring period, The value range of is [1, Nt+Yt], Nt represents the total number of synchronous change cycles, and Yt represents the total number of asynchronous change cycles; By formula: Obtain the cumulative first-order difference ratio Rw of the monitoring cycle-waste analysis index WI series W(t), where express The first-order difference is greater than 0, is the absolute value of the first-order difference; It should be noted that the cumulative first-order difference ratio measures the increasing trend of the entire sequence by comparing the sum of the increasing parts with the total change. If the cumulative first-order difference ratio Rw value is close to 1, it means that most of the first-order differences are positive numbers, that is, the increasing part of the sequence is dominant, and the overall trend is obvious. Compare the accumulated first-order difference ratio Rw with the difference threshold to determine whether the overall trend of the waste analysis index WI change curve of multiple monitoring cycles tends to increase or decrease; If the cumulative first-order difference ratio Rw is higher than the difference threshold, the overall trend of the waste analysis index WI change curve of multiple monitoring cycles tends to increase, otherwise, the overall trend of the waste analysis index WI change curve of multiple monitoring cycles tends to decrease; If the overall trend of the waste analysis index WI change curve of multiple monitoring cycles tends to decrease, no processing will be done; If the overall trend of the waste analysis index WI change curve of multiple monitoring cycles tends to increase, a waste over-limit warning is sent to the key analysis window; After sending the waste over-limit warning to the key analysis window, the key analysis window adjusts the amount of food supplied on the plate after receiving the waste over-limit warning sent by the window; It should be noted that the overall trend of the waste analysis index WI change curves for multiple monitoring cycles tends to increase, indicating that as the monitoring cycle progresses, the waste analysis index WI and the average value of the plates in the time period increase at the same time, that is, when the focus is on analyzing the increased customer demand at the window, the supply of dishes or the quality of service did not keep up, resulting in more food waste; Based on the waste over-limit warning, obtain the key analysis window, and the proportion of plate waste in the entire monitoring period in each analysis period P jk ; The proportion of plate waste in each analysis period during the entire monitoring period is P jk Perform summation and average processing, arrange the calculation results in order from large to small, and obtain the time period with the highest waste; Based on the highest waste period, obtain the plate waste amount W of the highest waste period in the entire monitoring cycle jk The mean ; Based on the plate waste amount W of the highest waste period in the entire monitoring period jk The mean , calculate the plate waste amount W during the highest waste period jk Mean coefficient of variation ; The amount of plate waste W jk The mean The amount of plate waste during the highest waste period W jk Mean coefficient of variation Perform product processing to obtain the waste control amount during the time period with the highest waste in the key analysis window; For the time period with the highest waste in the key analysis window, the food supply of each plate is regulated and reduced according to the waste control amount; If the waste analysis index WI of the key analysis window after adjustment is reduced, so that the original key analysis window is no longer the key analysis window, the adjustment is considered successful. Otherwise, the food supply of each plate in the key analysis window is further adjusted and reduced until the original key analysis window is no longer the key analysis window. Specifically, the way to further regulate and reduce the amount of food supplied to each plate in the focus analysis window is as follows: Recalculate the amount of plate waste W after regulation and reduction jk The mean , the amount of plate waste during the highest waste period W jk Mean coefficient of variation , and calculate the waste control amount of continued control; According to the waste control amount that continues to be regulated, the food supply of each plate in the key analysis window is further regulated and reduced; After obtaining the regulation reduction, the key analysis window is re-analyzed, and the food supply of each plate in the key analysis window is continued to be regulated to reduce the waste in the cafeteria and realize the intelligent management of the cafeteria waste phenomenon.

[0018] Embodiment 3 like Figure 2 As shown, a smart canteen management system includes the following modules: Data acquisition module: obtains the plate data from the canteen supply window in the canteen and establishes a plate waste data group; Data reconstruction module: Based on the plate waste data set, build a plate classification model, determine and analyze the waste data of the plate, remove invalid data in the plate waste data set, and build a waste analysis table X ij ; Window analysis module: Based on waste analysis table X ij , establish a waste analysis index WI analysis model, sort and analyze the waste analysis index WI of each canteen supply window, and determine the key analysis window; Trend analysis module: Based on the key analysis window, numerical analysis is performed on the number of plates in the monitoring period to obtain the plate value for the period. Curve analysis is performed on the waste analysis index and plate value for the period of multiple monitoring periods to determine whether the changing trends of the curves are consistent. Intelligent control module: If the waste analysis index and time period plate values ​​of multiple monitoring cycles have the same curve change trend, determine whether the waste analysis index change shows an increasing trend. If it is increasing, adjust the food supply of each plate in the key analysis window to achieve intelligent management of canteen waste.

[0019] Embodiment 4 Reference Figure 3The embodiment of the present invention further provides a computer device 3, comprising: a memory 302 and a processor 301 and a computer program 303 stored in the memory 302. When the computer program 303 is executed on the processor 301, a smart canteen management method as described in any one of the above methods is implemented.

[0020] The computer device 3 may be a computing device such as a desktop computer, a notebook, a palm computer, a cloud server, etc. The computer device 3 may include, but is not limited to, a processor 301 and a memory 302, which can be understood by those skilled in the art.

[0021] Figure 3 It is only an example of computer device 3 and does not constitute a limitation on computer device 3. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components, for example, it may also include input and output devices, network access devices, etc.

[0022] The processor 301 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0023] In some embodiments, the memory 302 may be an internal storage unit of the computer device 3, such as a hard disk or memory of the computer device 3. In other embodiments, the memory 302 may also be an external storage device of the computer device 3, such as a plug-in hard disk, a smart memory card (SmartMediaCard, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the computer device 3. Further, the memory 302 may also include both an internal storage unit of the computer device 3 and an external storage device. The memory 302 is used to store an operating system, an application program, a boot loader (BootLoader), data, and other programs, such as the program code of the computer program, etc. The memory 302 may also be used to temporarily store data that has been output or is to be output.

[0024] Embodiment 5 An embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, a smart canteen management method as described in any one of the above methods is implemented.

[0025] In this embodiment, if the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of the above-mentioned various method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium can at least include: any entity or device that can carry the computer program code to the camera / terminal device, recording medium, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium. For example, USB flash drive, mobile hard disk, disk or optical disk. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electric carrier signals and telecommunication signals.

[0026] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0027] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0028] In the embodiments disclosed in the present application, it should be understood that the disclosed devices / terminal equipment and methods can be implemented in other ways. For example, the device / terminal equipment embodiments described above are only schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0029] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0030] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0031] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A smart canteen management method, characterized in that: The following steps are involved: Step 1: Obtain the meal tray data from the cafeteria supply window in the cafeteria and establish a meal tray waste data group; Step 2: Based on the plate waste data set, build a plate classification model, remove invalid data in the plate waste data set, and build a waste analysis table X ij ; Step 3: Based on the waste analysis table X ij , establish a waste analysis index analysis model and determine the key analysis window; Step 4: Based on the key analysis window, numerical analysis is performed on the number of plates in the monitoring period to obtain the plate value for the period, and curve analysis is performed on the waste analysis index and plate value for the period of multiple monitoring periods to determine whether the change trend of the curve is consistent; Step 5. If the changing trend of the curve is consistent, determine whether the change in the waste analysis index shows an increasing trend. If it is increasing, adjust the food supply of each plate in the key analysis window to achieve intelligent management of cafeteria waste.

2. A smart canteen management method according to claim 1, characterized in that: The waste analysis table X ij The construction method is: The weight of the plates at the canteen supply window and the plate collection window is subtracted to obtain the amount of food waste; Obtain images of plates from the canteen supply window and plate recycling window, and use the twin neural network algorithm to build a plate classification model to determine whether the wasted food on the plates comes from the same supply window of the canteen; If the food wasted on the plate does not come from the same canteen supply window, the plate is marked as a plate that cannot be analyzed, and the amount of food wasted on the plate that cannot be analyzed is removed from the plate waste data group to construct a waste analysis data group; Obtain the waste analysis data set of all canteen supply windows in the canteen and construct the waste analysis table X ij , i is the number of the plate waste data group, and the value range of i is [1,n j ], j represents the number of the canteen supply window, and the value range of j is [1,m], n j Represents the total number of data in each cafeteria supply window waste analysis data group.

3. A smart canteen management method according to claim 1, characterized in that: The method for obtaining the key analysis window is as follows: Through the waste analysis index WI analysis model, the window waste index WI of each canteen supply window is obtained 3j , the fluctuation waste index WI of each canteen supply window 4j ; The window waste index WI 3j , Fluctuation waste index WI 4j Perform product processing to obtain the waste analysis index WI of each canteen supply window; Obtain the waste analysis index WI of each canteen supply window, arrange the waste analysis index WI of each canteen supply window in descending order, and determine the canteen supply window with the largest waste analysis index WI value as the key analysis window.

4. A smart canteen management method according to claim 3, characterized in that: The window waste index WI of each canteen supply window 3j The way to obtain is: By formula: Get the coefficient of variation CW of each canteen supply window j , where AW j is the average waste amount of the waste analysis data set within each cafeteria supply window; The sub-index WI1 based on the average waste amount and dispersion degree is established by the formula: , get the sub-index WI of the average waste and dispersion degree of each canteen supply window 1j ; Waste Analysis Table X ij According to the different food supply periods in the cafeteria, multiple analysis periods k are divided, k=1, 2, 3, 4, 5; Obtain the amount of wasted plates in the same analysis period of the canteen supply window and sum them up to obtain the amount of wasted plates in each analysis period W jk , j is the number of the canteen supply window, and k is the analysis period; By formula: Get the proportion of plate waste in each analysis period P jk ; By formula: Establish sub-index WI based on analysis period 2j , where w k is the proportionality factor for each analysis period; Sub-index WI based on meal period 2j , sub-index WI based on average waste and dispersion 1j , through weighted summation calculation, we get the window waste index WI 3j .

5. A smart canteen management method according to claim 3, characterized in that: The fluctuation waste index WI of each canteen supply window 4j The way to obtain is: through the formula: Get the fluctuation waste index WI 4j , where X ij max、X ij min are waste analysis table X ij The maximum and minimum values ​​of plate waste at the same cafeteria supply window.

6. A smart canteen management method according to claim 1, characterized in that: The method to judge whether the changing trend of the curve is consistent is: Conduct curve analysis on the waste analysis index and time period plate value of multiple monitoring cycles to obtain the phase difference of synchronous change cycles , the phase difference of the asynchronous change cycle , calculate the synchronous trend value Tb; If the synchronization trend value Tb is higher than the synchronization trend threshold, it indicates that the change trends of the waste analysis index and the time period plate value curves of multiple monitoring cycles are consistent.

7. A smart canteen management method according to claim 6, characterized in that: The method of performing curve analysis on the waste analysis index and time period plate value of multiple monitoring cycles is as follows: Based on the key analysis window, the waste analysis index WI in multiple monitoring cycles is obtained, and the monitoring cycle-waste analysis index WI sequence W(t) is constructed; Obtain the number of plates in each analysis period of the key analysis window within each monitoring cycle, sum and average the number of plates in each analysis period within the monitoring cycle, and obtain the plate value for the period; Obtain meal plate values ​​for multiple monitoring periods and construct a monitoring period-period meal plate value sequence S(t); In the two-dimensional rectangular coordinate system, the change curves of the monitoring cycle-waste analysis index WI and the monitoring cycle-time period plate average are drawn; The Hilbert-Huang transform is performed on the monitoring cycle-waste analysis index WI sequence W(t) and the monitoring cycle-time plate value sequence S(t) to construct a plate change analysis model; Based on the plate change analysis model, the phase difference of the synchronous change cycle is obtained , the phase difference of the asynchronous change cycle .

8. A smart canteen management method according to claim 1, characterized in that: The smart management method for canteen waste is: Based on the highest waste period, the plate waste amount W jk The mean The amount of plate waste during the highest waste period W jk Mean coefficient of variation Perform product processing to obtain the waste control amount during the time period with the highest waste in the key analysis window; For the time periods with the highest waste in the key analysis window, reduce the amount of food provided on each plate according to the waste control amount; If the waste analysis index WI does not decrease in the key analysis window after adjustment, the food supply amount of each plate in the key analysis window will be further adjusted and reduced until the original key analysis window is no longer a key analysis window.

9. A smart canteen management method according to claim 8, characterized in that: The method for judging whether the waste analysis index change shows an increasing trend is as follows: If the curve change trends of the waste analysis index WI and the time period meal plate value are consistent in multiple monitoring cycles; The cumulative first-order difference ratio Rw of the monitoring cycle-waste analysis index WI sequence W(t) is obtained through the first-order difference ratio calculation formula; If the accumulated first-order difference ratio Rw is higher than the difference threshold, the overall trend of the waste analysis index WI change curve of multiple monitoring cycles tends to increase.

10. A smart canteen management system, used to implement a smart canteen management method according to any one of claims 1 to 9, characterized in that: Includes the following modules: Data acquisition module: obtains the meal tray data from the canteen supply window in the canteen and establishes a meal tray waste data group; Data reconstruction module: Based on the plate waste data set, build a plate classification model, remove invalid data in the plate waste data set, and build a waste analysis table X ij ; Window analysis module: Based on waste analysis table X ij , establish a waste analysis index analysis model and determine the key analysis window; Trend analysis module: Based on the key analysis window, numerical analysis is performed on the number of plates in the monitoring period to obtain the plate value for the period. Curve analysis is performed on the waste analysis index and plate value for the period of multiple monitoring periods to determine whether the changing trends of the curves are consistent. Intelligent control module: If the changing trend of the curve is consistent, determine whether the change in the waste analysis index shows an increasing trend. If it is increasing, adjust the food supply of each plate in the key analysis window to achieve intelligent management of canteen waste.

Images