A sensor data monitoring method for intelligent material spreading and steaming process
By performing temperature monitoring and infrared image processing on the inner surface of the steamer, combining the characteristics of the high-temperature point set area and the differences in the inter-frame characteristics, a complete high-temperature steam area is selected, which solves the problem affected by the steamer edge effect in the existing technology, and realizes the Accurate identification and effective sprinkling of high-temperature steam areas.
Patent Information
- Application Number
- CN202411230805.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-04
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2044-09-04
AI Technical Summary
When using the K-means clustering algorithm in the prior art, it is difficult to obtain a complete high-temperature steam area due to the influence of the retort edge effect, resulting in data division falling into a local optimal solution and the high-temperature steam area cannot be accurately identified.
By monitoring the temperature of the inner surface of the steamer, a continuous multi-frame infrared image is obtained, clustered and screened high-temperature data clusters. Combined with the data point distribution characteristics and inter-frame feature differences in the high-temperature point set area, the suspected steam area is selected and finally determined Spread the steam area.
It achieves the complete high-temperature steam area not affected by the retort edge effect, improves the accuracy and aggregation of steam area identification, and ensures the efficiency and quality of the sprinkler process.
Smart Images

Figure CN118781374B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of data processing, and in particular to a sensor data monitoring method for an intelligent material spreading and steaming process. Background Art
[0002] The intelligent material spreading device is an advanced equipment in the brewing industry. It is mainly used to spread the raw materials evenly on the retort (still) during the brewing process to ensure the uniform distribution of heat and water during the distillation process. The use of this equipment helps to improve the efficiency and quality of brewing. In the process of spreading materials on the retort, it is very important to sense the steam distribution in the retort in real time and spread materials in the places where steam is emitted. The intelligent material spreading device monitors the temperature data in the retort through infrared monitoring sensors, obtains the approximate distribution area of steam based on the temperature data in the retort, and guides the robotic arm to spread materials.
[0003] The existing technology usually uses K-means clustering algorithm to process infrared temperature data to obtain areas with high temperature clustering characteristics as high temperature steam areas. However, the existing technology directly uses temperature difference and Euclidean distance as the allocation conditions for point clusters to cluster centers during the clustering process, which will be affected by the steamer edge effect (steam appears preferentially at the edge of the steamer) during the process of spreading materials on the steamer, causing the data division to fall into a local optimal solution and fail to obtain a complete high temperature steam area.
[0004] Therefore, how to obtain a complete high-temperature steam region with good aggregation characteristics becomes an urgent problem to be solved. Summary of the invention
[0005] In view of this, an embodiment of the present invention provides a sensor data monitoring method for an intelligent material spreading and steaming process to solve the problem of how to obtain a complete high-temperature steam area with good aggregation characteristics.
[0006] In an embodiment of the present invention, a sensor data monitoring method for an intelligent material spreading and retorting process is provided, and the method comprises the following steps:
[0007] The temperature of the inner surface of the steamer is monitored to obtain a plurality of continuous infrared images within a preset material spreading period, wherein the value of the data point in each frame of the infrared image is the temperature value;
[0008] For any frame of infrared image, cluster all data points in the infrared image to obtain at least two data clusters, select a high-temperature data cluster from all the data clusters, and obtain at least one high-temperature point set region in the infrared image based on the data points included in the high-temperature data cluster;
[0009] According to the data point distribution characteristics in each of the high-temperature point set areas, the degree of high-temperature steam in each of the high-temperature point set areas is obtained, and according to the degree of high-temperature steam in each of the high-temperature point set areas, at least one suspected steam area is screened out in all the high-temperature point set areas;
[0010] According to the difference in data point distribution characteristics between the suspected steam regions in the infrared images of adjacent frames, respectively obtain the degree of inter-frame feature difference of each of the suspected steam regions in the last frame of the infrared image, and according to the degree of inter-frame feature difference of each of the suspected steam regions, obtain the steam region in the last frame of the infrared image;
[0011] The steam area is sprinkled with material, and after the material is sprinkled, the steam area within the next preset spreading cycle is continuously acquired until the steamer is full and the process stops.
[0012] Preferably, obtaining the high temperature steam degree in each of the high temperature point set regions according to the data point distribution characteristics in each of the high temperature point set regions comprises:
[0013] For any high-temperature point set region, the number of data points in the high-temperature point set region is obtained, the number of data points is normalized to obtain a corresponding first normalized value, the data mean and data variance of the data points in the high-temperature point set region are calculated, the data mean is normalized to obtain a corresponding second normalized value, and the data variance is inversely normalized to obtain a corresponding third normalized value;
[0014] An average value among the first normalized value, the second normalized value, and the third normalized value is calculated as the high temperature steam degree in the high temperature point set region.
[0015] Preferably, the step of screening out at least one suspected steam region from all high temperature point set regions according to the degree of high temperature steam in each high temperature point set region comprises:
[0016] A preset high-temperature steam level threshold is obtained, and if the high-temperature steam level in any high-temperature point set region is greater than or equal to the high-temperature steam level threshold, the high-temperature point set region is regarded as a suspected steam region.
[0017] Preferably, the step of obtaining the inter-frame feature difference degree of each of the suspected steam regions in the last frame of the infrared image according to the difference in data point distribution features between the suspected steam regions in the infrared images of adjacent frames comprises:
[0018] For any suspected steam region in the last frame of the infrared image, obtain the suspected steam region of all frames within the preset material spreading cycle to form a set, and obtain the maximum suspected steam region in each frame of the infrared image except the suspected steam region in the set;
[0019] According to the acquisition order of all infrared images, all the maximum suspected steam regions and the suspected steam regions are sorted to obtain a suspected steam region sequence, first absolute values of the difference in the number of data points between every two adjacent suspected steam regions in the suspected steam region sequence are calculated to obtain an accumulated value of the first absolute values of the difference, and the accumulated value of the first absolute values of the difference is normalized to obtain a corresponding fourth normalized value;
[0020] Respectively calculating the second difference absolute values of the data point means between every two adjacent suspected steam regions in the suspected steam region sequence to obtain an accumulated value of the second difference absolute values, and normalizing the accumulated value of the second difference absolute values to obtain a corresponding fifth normalized value;
[0021] The sum of the fourth normalized value and the fifth normalized value is calculated as the inter-frame feature difference degree of the suspected steam region in the last frame of the infrared image.
[0022] Preferably, obtaining the steam region in the last frame of the infrared image according to the inter-frame feature difference degree of each of the suspected steam regions includes:
[0023] A preset inter-frame feature difference degree threshold is obtained. If the inter-frame feature difference degree of any suspected steam region is greater than or equal to the inter-frame feature difference degree threshold, the suspected steam region is used as the steam region in the last frame of the infrared image.
[0024] Preferably, the step of screening all data clusters to obtain a high temperature data cluster includes:
[0025] According to all the data points in each of the data clusters, a data mean of the data points in each of the data clusters is obtained, and a data cluster corresponding to the largest data mean is selected as a high-temperature data cluster.
[0026] Preferably, acquiring at least one high-temperature point set region in the infrared image according to the data points included in the high-temperature data cluster comprises:
[0027] Each data point in the high temperature data cluster is used as a high temperature point, and at least one high temperature point set region is obtained according to the positions of all high temperature points in the infrared image.
[0028] Preferably, the temperature of the inner surface of the steamer pot is monitored to obtain a continuous multi-frame infrared image within a preset material spreading period, including:
[0029] An infrared temperature sensor is installed above the steamer pot to collect infrared images of the inner surface of the steamer pot at a preset sampling frequency to obtain continuous multi-frame infrared images within a preset material spreading cycle.
[0030] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:
[0031] The present invention monitors the temperature of the inner surface of the steamer pot to obtain continuous multiple frames of infrared images within a preset material spreading cycle, wherein the value of the data point in each frame of the infrared image is a temperature value; for any frame of infrared image, cluster all data points in the infrared image to obtain at least two data clusters, filter out a high-temperature data cluster from all data clusters, and obtain at least one high-temperature point set area in the infrared image according to the data points included in the high-temperature data cluster; obtain the degree of high-temperature steam in each high-temperature point set area according to the distribution characteristics of the data points in each high-temperature point set area, and filter out at least one suspected steam area in all high-temperature point set areas according to the degree of high-temperature steam in each high-temperature point set area; obtain the degree of inter-frame feature difference of each suspected steam area in the last frame of the infrared image according to the difference in data point distribution characteristics between the suspected steam areas in the infrared images of adjacent frames, and obtain the steam area in the last frame of the infrared image according to the degree of inter-frame feature difference of each suspected steam area; spread the material in the steam area, and continue to obtain the steam area in the next preset material spreading cycle after spreading the material until the steamer pot is full and stops. The present invention clusters the collected data in the steamer, obtains the high-temperature point set area according to the clustering result, obtains the suspected steam area in combination with the data distribution characteristics in the high-temperature point set area, and further obtains the steam area in combination with the difference in data point distribution characteristics between the suspected steam areas. On the basis of ensuring that the clustering results have both temperature clustering and steam characteristic clustering, a complete clustering result that is not affected by the steamer edge effect is obtained, so that finally a complete high-temperature steam area with good clustering characteristics is obtained. BRIEF DESCRIPTION OF THE DRAWINGS
[0032] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. 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 paying creative labor.
[0033] Figure 1 It is a method flow chart of a sensor data monitoring method for an intelligent material spreading and steaming process provided by Embodiment 1 of the present invention;
[0034] Figure 2 It is a K-means clustering result affected by the retort inner and outer effects provided by an embodiment of the present invention;
[0035] Figure 3 An intelligent material spreading and steaming robot arm device is provided in an embodiment of the present invention. DETAILED DESCRIPTION
[0036] Embodiments of the present disclosure are described in detail below, and examples of the embodiments are shown in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present disclosure, but should not be understood as limiting the present disclosure.
[0037] It should be noted that the terms "first", "second", etc. in the specification of the present disclosure and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the present disclosure.
[0038] In order to illustrate the technical solution of the present invention, specific embodiments are provided below for illustration.
[0039] See also Figure 1 , is a method flow chart of a sensor data monitoring method for an intelligent material spreading and retorting process provided in Embodiment 1 of the present invention, such as Figure 1 As shown, the method may include:
[0040] Step S101, monitoring the temperature of the inner surface of the steamer pot to obtain a plurality of continuous infrared images within a preset material spreading cycle, wherein the value of a data point in each frame of the infrared image is a temperature value.
[0041] An infrared temperature sensor is installed above the steamer pot to collect temperature data of the inner surface of the steamer pot. The preset collection frequency is once per second. Five consecutive infrared images of the steamer material surface inside the steamer pot are collected as a preset material spreading cycle. The infrared images collected per second are regarded as one frame. There is no limitation here. According to the specific implementation scenario setting, each pixel in the infrared image can be used as a data point, and the value of each data point is the temperature value.
[0042] Step S102, for any frame of infrared image, cluster all data points in the infrared image to obtain at least two data clusters, screen all data clusters to obtain a high-temperature data cluster, and obtain at least one high-temperature point set area in the infrared image based on the data points included in the high-temperature data cluster.
[0043] When K-means clustering is performed on the data points in the infrared image monitored by the infrared temperature sensor, due to the effect of the inner and outer edges of the retort, when the temperature is concentrated at the edge of the retort, the data partition is prone to fall into a local optimal solution and the complete high-temperature steam area at the edge of the retort cannot be obtained. Figure 2 As shown, a K-means clustering result affected by the steamer inner and steamer edge effect provided by an embodiment of the present invention is provided.
[0044] Therefore, in the embodiment of the present invention, firstly, for the data points in each frame of the infrared image, all the data points in the infrared image are clustered by K-means to obtain multiple data clusters, wherein the K value in the K-means clustering is obtained by the elbow method, and the K-means clustering and the elbow method belong to the prior art and will not be described here. Because the temperature of the steam is much higher than the surface temperature of the steamer material, the clustering results obtained by the numerical clustering of the data points in the infrared image will divide the high-temperature data and other low-temperature data into different data clusters, and then it is necessary to screen all the data clusters to obtain the high-temperature data cluster.
[0045] Preferably, screening all data clusters to obtain a high-temperature data cluster includes: obtaining a data mean of the data points in each data cluster based on all data points in each data cluster, and selecting a data cluster corresponding to the largest data mean as the high-temperature data cluster.
[0046] In one embodiment, taking the jth data cluster as an example, the temperature mean is calculated based on the temperature value corresponding to each data point in the jth data cluster. Similarly, the temperature mean corresponding to each data cluster is obtained, and the data cluster corresponding to the maximum temperature mean is selected as the high-temperature data cluster.
[0047] Furthermore, after obtaining the high temperature data cluster, at least one high temperature point set region is obtained in the infrared image according to the data points included in the high temperature data cluster. The method for obtaining the high temperature point set region is:
[0048] Each data point in the high temperature data cluster is taken as a high temperature point, and at least one high temperature point set region is obtained according to the positions of all high temperature points in the infrared image.
[0049] In one embodiment, each data point in the high temperature data cluster is regarded as a high temperature point, and the high temperature point is mapped to the infrared image. According to the position of the high temperature point in the infrared image, the high temperature points in the infrared image are divided into different areas, and each area is a high temperature point set area.
[0050] At this point, all high-temperature point set areas in each frame of infrared image can be obtained.
[0051] Step S103, obtaining the degree of high-temperature steam in each of the high-temperature point set areas according to the data point distribution characteristics in each of the high-temperature point set areas, and screening out at least one suspected steam area in all the high-temperature point set areas according to the degree of high-temperature steam in each of the high-temperature point set areas.
[0052] Because there are many gaps in the retort material, steam may come out from anywhere, so there will be many small gaps on the surface of the retort material, resulting in small steam areas. These steam areas cannot form a complete steam area and are not suitable for spreading materials. However, they will interfere with the recognition of the steam area by the spreading robot arm. Therefore, we cannot rely solely on obtaining the high-temperature point set area as the steam area for spreading materials. We need to obtain the high-temperature point set area with clustering based on the spatial information of the data points in the infrared image. If clustering based on Euclidean distance is used directly, it will be affected by the retort edge effect, resulting in the high-temperature point set area being locally clustered and unable to represent the complete steam area. Therefore, it is necessary to obtain the suspected steam area with strong clustering based on the data distribution characteristics of the data points in the high-temperature point set area.
[0053] Among them, the method of obtaining the suspected steam area with strong aggregation according to the data distribution characteristics of the data points in the high-temperature point set area is as follows:
[0054] (1) According to the distribution characteristics of the data points in each high-temperature point set area, the high-temperature steam degree in each high-temperature point set area is obtained.
[0055] Specifically, for any high-temperature point set area, the number of data points in the high-temperature point set area is obtained, the number of data points is normalized to obtain a corresponding first normalized value, the data mean and data variance of the data points in the high-temperature point set area are calculated, the data mean is normalized to obtain a corresponding second normalized value, the data variance is inversely normalized to obtain a corresponding third normalized value; and the mean of the first normalized value, the second normalized value and the third normalized value is calculated as the high-temperature steam degree in the high-temperature point set area.
[0056] In one embodiment, the degree of high-temperature steam is comprehensively obtained by the number of data points in the high-temperature point set area, the data mean of the data points, and the data uniformity of the data points. The more data points in the high-temperature point set area and the stronger the data point aggregation, the more consistent with the steam area characteristics; because the heat molecules in the large steam area conduct heat to each other and the heating rate is faster, the temperature of the large steam area is higher than the average temperature level of the small gap steam area, so the more data points in the high-temperature point set area, the stronger the steam area characteristics; the higher the data mean in the high-temperature point set area, the stronger the steam area characteristics; in the steaming process, since steam has good thermal conductivity, it can quickly and evenly distribute heat to maintain a relatively constant temperature, so the temperature of the large steam area is usually more stable, while the steam area with small gaps is affected by the surface temperature of the surrounding steam material, and the uniformity of multiple data points on the internal boundary of the area is poor. Based on the small overall area range of the steam area with small gaps, the temperature uniformity of the steam area with small gaps is poor, so the stronger the data point aggregation in the high-temperature point set area, the stronger the steam area characteristics. Based on the above three characteristics, the high-temperature steam degree of the high-temperature point collection area is obtained, and the steam characteristics of each high-point temperature collection area are evaluated.
[0057] Taking the pth high-temperature point set area as an example, calculate its high-temperature steam degree:
[0058]
[0059] in: is the high temperature steam degree of the pth high temperature point set area; p is the sequence number of the high temperature point set area; is the number of data points in the pth high temperature point set area; i is the sequence number of the data point in the high temperature point set area; is the value of the i-th data point in the p-th high temperature point set area; is the data mean of all data points in the pth high temperature point set area; Represents the variance of the values of all data points in the pth high temperature point set area; is the inverse normalization function; is a linear normalization function.
[0060] It should be noted that for the pth high temperature point set area, the number of data points in the pth high temperature point set area The more it is, the more it conforms to the characteristics of the steam area, and the stronger the high-temperature steam is; the data mean of all data points in the pth high-temperature point set area The higher the value, the stronger the high-temperature steam level; the variance of the values of all data points in the pth high-temperature point set area The smaller it is, the stronger the uniformity within the pth high-temperature point set area, the better the stability, and the stronger the degree of high-temperature steam; the stronger the degree of high-temperature steam in the pth high-temperature point set area, the more likely the pth high-temperature point set area is a steam area.
[0061] (2) According to the degree of high-temperature steam in each high-temperature point set area, at least one suspected steam area is screened out from all high-temperature point set areas, including:
[0062] A preset high-temperature steam level threshold is obtained, and if the high-temperature steam level in any high-temperature point set region is greater than or equal to the high-temperature steam level threshold, the high-temperature point set region is regarded as a suspected steam region.
[0063] In one embodiment, the high temperature steam level threshold is set to 0.6, which is not limited here and can be set according to the implementation scenario. When the high temperature steam level in the pth high temperature point set area is When , the pth high-temperature point set region is considered to be a suspected steam region, thereby obtaining the suspected steam region in each frame of infrared image.
[0064] At this point, all suspected steam regions in each frame of infrared image can be obtained.
[0065] Step S104, according to the difference in data point distribution characteristics between the suspected steam areas in the infrared images of adjacent frames, respectively obtain the degree of inter-frame feature difference of each of the suspected steam areas in the infrared image of the last frame, and according to the degree of inter-frame feature difference of each of the suspected steam areas, obtain the steam area in the infrared image of the last frame.
[0066] During the process of spreading the material onto the steamer, the steamer material may be highly susceptible to heat. In some areas, the bottom of the steamer material is exposed to a large amount of steam heat and heats up, but no steam is emitted yet. Under the influence of the temperature of the steamer material at this time, this area is easily detected as a high-temperature point cluster area, and the clustering is strong, which can easily lead to false detection. The high-temperature point cluster area affected by the steamer material is identified as a steam area. For this, it is necessary to combine the inter-frame feature differences between infrared images collected in multiple frames to obtain the real steam area.
[0067] The generation of steam has a certain process. From the beginning of steaming to regional steaming, the range of the steam area gradually increases, the temperature gradually rises, and the temperature increases rapidly. However, due to the poor thermal conductivity of the retort material, the range of the high-temperature point set area affected by the retort material is low and the temperature rises slowly. After the last material spreading, the temperature and range of the area where steaming will be generated are gradually expanded. The steam area in the latest infrared image must include the steam area at this position in the previous infrared image. Therefore, the last frame of infrared image is used as the analysis object. For each suspected steam area, the maximum suspected steam area in the suspected steam area in the last infrared image and the suspected steam area in the previous infrared image is obtained, and the range and temperature difference between the two is obtained as the inter-frame feature difference. All inter-frame feature differences in the material spreading cycle are iteratively obtained. The overall inter-frame feature difference is large, indicating that the corresponding suspected steam area is more likely to be a real steam area, thereby eliminating the risk of false detection of the high-temperature retort material area, and further improving the monitoring accuracy of the high-temperature steam area. Therefore, it is necessary to obtain the steam area based on the inter-frame feature difference between adjacent frames in the suspected steam area.
[0068] The method for obtaining the steam region according to the inter-frame feature difference between adjacent frames in the suspected steam region is as follows:
[0069] (1) According to the difference in data point distribution characteristics between the suspected steam regions in the infrared images of adjacent frames, the degree of inter-frame characteristic difference of each of the suspected steam regions in the last frame of the infrared image is obtained respectively.
[0070] Specifically, for any suspected steam area in the last frame of the infrared image, obtain the suspected steam areas of all frames of the suspected steam area within the preset spreading cycle to form a set, and obtain the maximum suspected steam area in each frame of the infrared image except the suspected steam area in the set; sort all the maximum suspected steam areas and the suspected steam areas according to the acquisition order of all infrared images to obtain a suspected steam area sequence, calculate the first absolute difference value of the number of data points between every two adjacent suspected steam areas in the suspected steam area sequence, obtain the accumulated value of the first absolute difference value, normalize the accumulated value of the first absolute difference value, and obtain the corresponding fourth normalized value; calculate the second absolute difference value of the mean of the data points between every two adjacent suspected steam areas in the suspected steam area sequence, obtain the accumulated value of the second absolute difference value, normalize the accumulated value of the second absolute difference value, and obtain the corresponding fifth normalized value; calculate the sum of the fourth normalized value and the fifth normalized value as the inter-frame feature difference degree of the suspected steam area in the last frame of the infrared image.
[0071] In one embodiment, taking the qth suspected steam area in the last frame of infrared image in the material spreading cycle as an example, the suspected steam areas of the qth suspected steam area in all frames in the preset material spreading cycle are obtained to form a set, and the maximum suspected steam area in each frame of infrared image except the qth suspected steam area is obtained in the set. For example, taking 5 frames of infrared images as a material spreading cycle, there is no limitation here, and it can be set according to the specific implementation scenario. For the qth suspected steam area in the 5th frame, the suspected steam areas in the first 4 frames contained in the range corresponding to the qth suspected steam area and the qth suspected steam area are obtained to form a set. In the set, the first frame of infrared image includes n suspected steam areas, and the one with the largest area range among the n suspected steam areas is selected as the maximum suspected steam area of the first frame of infrared image. Similarly, the maximum suspected steam area in each frame of infrared image within the range corresponding to the qth suspected steam area in the material spreading cycle can be obtained.
[0072] Taking the qth suspected steam region as an example, calculate the degree of feature difference between frames:
[0073]
[0074] in, is the degree of inter-frame feature difference of the qth suspected steam region in the last frame of infrared image of the spreading cycle; n is the number of frames in the spreading cycle; is the number of data points of the largest suspected steam area corresponding to the qth suspected steam area in the last frame of infrared image in the i-th frame of infrared image; i is the sequence number of each frame of infrared image in the spreading cycle; is the data point average of the largest suspected steam area corresponding to the qth suspected steam area in the last frame of infrared image in the i-th frame of infrared image; is a linear normalization function, and || represents an absolute value symbol. In this embodiment, the number of frames used is 5, and the number of inter-frames is 4, which can be adjusted according to scene requirements.
[0075] It should be noted that, for the qth suspected steam area in the last frame of infrared image in the spreading cycle, within the range corresponding to the qth suspected steam area, except for the last frame of infrared image, the largest suspected steam area in each of the previous frames of infrared image is the largest suspected steam area in each of the previous frames of infrared image. It is the absolute value of the difference in the number of data points of the suspected steam area in the previous i-th frame and the previous i-1-th frame in the maximum suspected steam area sequence corresponding to the q-th suspected steam area in the last frame of infrared image in the spreading cycle. When i=1, i-1=0 indicates the last frame of infrared image. The larger the difference, the larger the range difference of the q-th suspected steam area between the two frames. is the cumulative sum of the overall inter-frame differences of the qth suspected steam region, and the cumulative sum of the overall inter-frame differences The larger it is, the more likely the qth region is a true steam region. is the absolute value of the mean difference between the data points of the suspected steam area in the previous i-th frame and the previous i-1-th frame of the q-th suspected steam area in the last frame of infrared image during the spreading cycle. is the cumulative sum of the mean difference of the data in the qth suspected steam area. The larger it is, the faster the overall temperature changes, and the more likely the qth suspected steam area is to be a real steam area.
[0076] (2) According to the degree of difference in the inter-frame features of each suspected steam region, the steam region in the last frame of the infrared image is obtained, including:
[0077] A preset inter-frame feature difference degree threshold is obtained. If the inter-frame feature difference degree of any suspected steam region is greater than or equal to the inter-frame feature difference degree threshold, the suspected steam region is used as the steam region in the last frame of the infrared image.
[0078] In one embodiment, the inter-frame feature difference threshold is set to 0.8, which can be adjusted according to the actual experimental results of the scene. When q is the suspected steam region, it is considered that the qth suspected steam region is the steam region in the last frame of infrared image, thereby obtaining all steam regions in the last frame of infrared image in the spreading cycle.
[0079] At this point, all steam areas in the last frame of infrared image in the spreading cycle can be obtained.
[0080] Step S105, spreading the material in the steam area, and continuing to obtain the steam area in the next preset spreading cycle after spreading the material, until the steamer is full and stops.
[0081] like Figure 3 As shown, it is an intelligent material spreading and retorting mechanical arm device provided by an embodiment of the present invention. After real-time monitoring and obtaining the steam area, the mechanical arm spreads the material at the location of the steam area. After spreading the material, monitoring continues. If the steam area is not obtained within a monitoring and spreading cycle, the next spreading cycle is monitored until the steam area is obtained and the material is spread. Based on this rule, it is iterated repeatedly until the retort pot is filled after manual observation, and then the mechanical arm stops.
[0082] In summary, the present invention monitors the temperature of the inner surface of the steamer pot to obtain continuous multiple frames of infrared images within a preset material spreading cycle, wherein the numerical value of the data point in each frame of the infrared image is a temperature value; for any frame of infrared image, all data points in the infrared image are clustered to obtain at least two data clusters, and a high-temperature data cluster is screened out from all data clusters, and at least one high-temperature point set area is obtained in the infrared image according to the data points included in the high-temperature data cluster; according to the distribution characteristics of the data points in each of the high-temperature point set areas, the degree of high-temperature steam in each of the high-temperature point set areas is obtained, and according to the degree of high-temperature steam in each of the high-temperature point set areas, at least one suspected steam area is screened out from all high-temperature point set areas; according to the difference in data point distribution characteristics between the suspected steam areas in the infrared images of adjacent frames, the degree of inter-frame feature difference of each of the suspected steam areas in the last frame of the infrared image is obtained respectively, and according to the degree of inter-frame feature difference of each of the suspected steam areas, the steam area in the last frame of the infrared image is obtained; the material is spread in the steam area, and after spreading the material, the steam area in the next preset material spreading cycle continues to be obtained until the steamer pot is full and stops. The present invention clusters the collected data in the steamer, obtains the high-temperature point set area according to the clustering result, obtains the suspected steam area in combination with the data distribution characteristics in the high-temperature point set area, and further obtains the steam area in combination with the difference in data point distribution characteristics between the suspected steam areas. On the basis of ensuring that the clustering results have both temperature clustering and steam characteristic clustering, a complete clustering result that is not affected by the steamer edge effect is obtained, so that finally a complete high-temperature steam area with good clustering characteristics is obtained.
[0083] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention.
Claims
1. A sensor data monitoring method for an intelligent material spreading and retorting process, characterized in that: The method comprises: The temperature of the inner surface of the steamer is monitored to obtain a plurality of continuous infrared images within a preset material spreading period, wherein the value of the data point in each frame of the infrared image is the temperature value; For any frame of infrared image, cluster all data points in the infrared image to obtain at least two data clusters, select a high-temperature data cluster from all the data clusters, and obtain at least one high-temperature point set region in the infrared image based on the data points included in the high-temperature data cluster; According to the data point distribution characteristics in each of the high-temperature point set areas, the degree of high-temperature steam in each of the high-temperature point set areas is obtained, and according to the degree of high-temperature steam in each of the high-temperature point set areas, at least one suspected steam area is screened out in all the high-temperature point set areas; According to the difference in data point distribution characteristics between the suspected steam regions in the infrared images of adjacent frames, respectively obtain the degree of inter-frame feature difference of each of the suspected steam regions in the last frame of the infrared image, and according to the degree of inter-frame feature difference of each of the suspected steam regions, obtain the steam region in the last frame of the infrared image; Sprinkling the steam area with material, and continuing to obtain the steam area in the next preset spreading cycle after spreading the material, until the steamer is full and stops; The step of obtaining the degree of high temperature steam in each of the high temperature point set regions according to the distribution characteristics of the data points in each of the high temperature point set regions comprises: For any high-temperature point set region, the number of data points in the high-temperature point set region is obtained, the number of data points is normalized to obtain a corresponding first normalized value, the data mean and data variance of the data points in the high-temperature point set region are calculated, the data mean is normalized to obtain a corresponding second normalized value, and the data variance is inversely normalized to obtain a corresponding third normalized value; calculating the average of the first normalized value, the second normalized value and the third normalized value as the high temperature steam degree in the high temperature point set region; The step of obtaining the steam region in the last frame of the infrared image according to the inter-frame feature difference degree of each of the suspected steam regions comprises: A preset inter-frame feature difference degree threshold is obtained. If the inter-frame feature difference degree of any suspected steam region is greater than or equal to the inter-frame feature difference degree threshold, the suspected steam region is used as the steam region in the last frame of the infrared image.
2. The sensor data monitoring method for the intelligent material spreading and retorting process according to claim 1 is characterized in that: The step of screening all high temperature point set regions to obtain at least one suspected steam region according to the high temperature steam level in each high temperature point set region comprises: A preset high-temperature steam level threshold is obtained, and if the high-temperature steam level in any high-temperature point set region is greater than or equal to the high-temperature steam level threshold, the high-temperature point set region is regarded as a suspected steam region.
3. The sensor data monitoring method for the intelligent material spreading and retorting process according to claim 1 is characterized in that: The step of obtaining the inter-frame feature difference degree of each of the suspected steam regions in the last frame of the infrared image according to the difference in data point distribution features between the suspected steam regions in the infrared images of adjacent frames comprises: For any suspected steam region in the last frame of the infrared image, obtain the suspected steam region of all frames within the preset material spreading cycle to form a set, and obtain the maximum suspected steam region in each frame of the infrared image except the suspected steam region in the set; According to the acquisition order of all infrared images, all the maximum suspected steam regions and the suspected steam regions are sorted to obtain a suspected steam region sequence, first absolute values of the difference in the number of data points between every two adjacent suspected steam regions in the suspected steam region sequence are calculated to obtain an accumulated value of the first absolute values of the difference, and the accumulated value of the first absolute values of the difference is normalized to obtain a corresponding fourth normalized value; Respectively calculating the second difference absolute values of the data point means between every two adjacent suspected steam regions in the suspected steam region sequence to obtain an accumulated value of the second difference absolute values, and normalizing the accumulated value of the second difference absolute values to obtain a corresponding fifth normalized value; The sum of the fourth normalized value and the fifth normalized value is calculated as the inter-frame feature difference degree of the suspected steam region in the last frame of the infrared image.
4. The sensor data monitoring method for the intelligent material spreading and retorting process according to claim 1 is characterized in that: The high temperature data clusters are obtained by screening all the data clusters, including: According to all the data points in each of the data clusters, a data mean of the data points in each of the data clusters is obtained, and a data cluster corresponding to the largest data mean is selected as a high-temperature data cluster.
5. The sensor data monitoring method for the intelligent material spreading and retorting process according to claim 1 is characterized in that: The step of acquiring at least one high temperature point set region in the infrared image according to the data points included in the high temperature data cluster comprises: Each data point in the high temperature data cluster is taken as a high temperature point, and at least one high temperature point set region is obtained according to the positions of all high temperature points in the infrared image.
6. The sensor data monitoring method for the intelligent material spreading and retorting process according to claim 1 is characterized in that: The temperature of the inner surface of the steamer is monitored to obtain a continuous multi-frame infrared image within a preset material spreading cycle, including: An infrared temperature sensor is installed above the steamer pot to collect infrared images of the inner surface of the steamer pot at a preset sampling frequency to obtain continuous multi-frame infrared images within a preset material spreading cycle.
