Rice deep processing monitoring system and method based on Internet of Things

Abstract

This invention discloses an IoT-based monitoring system and method for rice deep processing, relating to the field of rice deep processing technology. It includes a data acquisition module, a data analysis module, a spatial calibration point acquisition module, a baseline topology coefficient acquisition module, and a real-time monitoring module. The real-time monitoring module acquires real-time data multiple times, analyzes whether monitored cadmium morphology feature points belong to the corresponding cadmium morphology distribution space, calculates the location qualification rate and monitoring topology coefficient, thereby determining whether the cadmium morphology distribution is abnormal and generating an abnormal distribution signal. This achieves real-time monitoring and accurate analysis of cadmium morphology during rice deep processing, enabling timely detection of cadmium content abnormalities, avoiding batch contamination, reducing resource waste, further ensuring the stability and safety of the processing process, and improving the quality of rice products.

Description

Technical Field

[0001] This invention belongs to the field of rice deep processing technology, specifically a monitoring system and method for rice deep processing based on the Internet of Things. Background Technology

[0002] Cadmium contamination in rice is one of the important issues in the field of food security. The processing precision has a significant impact on the residual amount and chemical form of cadmium in rice. Cadmium is a non-essential heavy metal element. Long-term intake can cause health problems such as kidney damage and osteoporosis. The enrichment of cadmium in rice mainly originates from soil pollution and is often concentrated in the rice husk, aleurone layer and germ. Moreover, its existence form in rice (such as inorganic cadmium, protein-bound cadmium, ester-soluble cadmium, etc.) has a decisive impact on bioavailability (toxicity). The chemical forms of cadmium are divided into free cadmium, protein-bound cadmium and organic acid chelated cadmium.

[0003] Existing methods for detecting cadmium content in rice are mostly offline laboratory tests, which are time-consuming and costly. Furthermore, these methods typically involve sampling and testing after product processing, making it impossible to achieve real-time monitoring and early warning during the processing. Once excessive levels are detected, mass contamination has often already occurred, leading to a significant waste of resources.

[0004] Furthermore, the bioavailability and toxicity of cadmium vary significantly in different forms (such as Cd2+, CdS, Cd(OH)2, etc.), and simply monitoring the total cadmium content cannot comprehensively assess its potential risks to the human body. However, current technologies struggle to achieve real-time online monitoring of multiple cadmium forms during rice processing, cannot accurately analyze the dynamic distribution characteristics of cadmium at different processing stages, and cannot accurately quantify and represent the dynamic distribution characteristics of cadmium forms in a multi-dimensional feature space. This results in the inability to intelligently determine abnormalities in processing stages, leaving rice quality control at a reactive, post-processing stage, hindering accurate risk assessment. Therefore, this paper proposes an IoT-based monitoring system and method for deep rice processing. Summary of the Invention

[0005] The purpose of this invention is to provide an IoT-based monitoring system and method for the deep processing of rice, which solves the technical problem that existing technologies are unable to achieve real-time online monitoring of multiple forms of cadmium during the rice processing process and cannot accurately analyze the dynamic distribution characteristics of cadmium in different processing stages.

[0006] An IoT-based monitoring system for the intensive processing of rice includes:

[0007] Step 1: Obtain historical content data of cadmium in various forms at each stage of rice processing;

[0008] Step 2: Analyze the historical content data of cadmium in each form at each stage of rice refining process to obtain the spatial distribution of cadmium in each refining stage.

[0009] Step 3: Based on the cadmium morphology characteristic points corresponding to each refining stage during multiple batches of rice processing, analyze and obtain the spatial calibration points corresponding to each refining stage;

[0010] Step 4: Analyze the cadmium morphological characteristic points and spatial calibration points corresponding to each batch of rice at each refining stage to obtain the reference topological coefficients corresponding to each refining stage.

[0011] Step 5: Acquire multiple cadmium morphology feature points of rice in the monitoring process, analyze whether each cadmium morphology feature point belongs to the cadmium morphology distribution space of the corresponding monitoring process, obtain the location qualification rate of the monitoring process, acquire the monitoring topology coefficient based on multiple cadmium morphology feature points, and determine the generation of abnormal distribution signals based on the location qualification rate of rice in the monitoring process and the monitoring topology coefficient.

[0012] As a further aspect of the present invention, the specific method for obtaining the cadmium morphology distribution space corresponding to each finishing stage is as follows:

[0013] One of the multiple refining stages is selected without replacement as the target stage. Historical cadmium content data of various forms at the target stage are obtained from multiple batches of rice processing. Historical cadmium content data Gi(GAi, GBi, GCI) of various forms of cadmium at the target stage are obtained from multiple batches of processing. The maximum values ​​GAmax, GBmax, and GCmax of different forms of cadmium GAi, GBi, and GCI are obtained from all batches. Gi(GAi, GBi, GCI) is normalized by GHi(GAi / GAmax, GBi / GBmax, GBi / GCmax) to obtain the cadmium morphology feature points GHi corresponding to each batch at the target stage. After discretization complex analysis of each cadmium morphology feature point GHi, the cadmium morphology distribution space Q1 corresponding to the target stage is generated, where i represents different processing batches, i = 1, 2, ..., n, n represents the total number of batches, and GAi, GBi, and GCI represent different forms of cadmium.

[0014] Using the same analytical method, the cadmium speciation distribution space Qb corresponding to each refining stage can be obtained, where b represents different refining stages, b = 1, 2, ..., r, and r represents the total number of refining stages.

[0015] As a further aspect of the present invention, the specific method for discretizing and performing complex analysis on each cadmium morphological feature point GHi is as follows:

[0016] Centered on the cadmium morphology feature point GHi, the mean Gp of the distance between all cadmium morphology feature points and the standard deviation of the distance between all batch feature points are obtained according to GHi(GAi / GAmax, GBi / GBmax, GBi / GCmax). The sum of the mean Gp and twice the standard deviation is used as the preset radius β, i.e., β = Gp + 2Gp. Spheres are drawn according to the preset radius β, and multiple spheres Ai are obtained to correspond to the complex Fi. The union of the geometric regions covered by the complex F corresponding to each sphere Ai is used as the cadmium morphology distribution space Q1 corresponding to the target link.

[0017] As a further aspect of the present invention, the specific method for obtaining the spatial calibration points corresponding to each finishing stage is as follows:

[0018] First, the mean values ​​of each cadmium morphology characteristic data GAi / GAmax, GBi / GBmax, and GBi / GCmax are obtained from the cadmium morphology characteristic points GHi (GAi / GAmax, GBi / GBmax, GBi / GCmax) corresponding to each batch in the target stage. Then, the spatial calibration point B1 (BA, BB, BC) corresponding to the target stage is obtained. Using the same analysis method, the spatial calibration point Bb corresponding to each finishing stage can be obtained.

[0019] As a further aspect of the present invention: obtaining the reference topology coefficients corresponding to each finishing stage, specifically in the following manner:

[0020] Obtain the cadmium morphology characteristic points GHi (GAi / GAmax, GBi / GBmax, GBi / GCmax) and the spatial calibration points B1 (BA, BB, BC) corresponding to each batch in the target stage.

[0021] The distance Di between each cadmium morphology feature point and the spatial calibration point is calculated using a distance formula. Based on a preset radius β, the cadmium morphology distribution space Q1 corresponding to the target stage is divided into a spherical neighborhood centered on the spatial calibration point B1. Cadmium morphology feature points with a distance Di less than or equal to the preset radius β are identified as feature points falling within the neighborhood of calibration point B1. The number e of feature points falling within the neighborhood of calibration point B1 is counted, and the number e is compared with the volume V of the neighborhood of calibration point B1. B1 The ratio between them is used as the neighborhood density P corresponding to the target element. B1 ; through T1=α×P B1+(1-α) to calculate the reference topology coefficient T1 corresponding to the target stage, where α ranges from 0.5 to 0.7; using the same analysis method as for obtaining the reference topology coefficient T1 corresponding to the target stage, the reference topology coefficient Tb corresponding to each finishing stage can be obtained.

[0022] As a further aspect of the present invention: obtaining the neighborhood volume V of calibration point B1. B1 The specific method is as follows:

[0023] pass Obtain the volume V of the neighborhood of calibration point B1 B1 .

[0024] As a further aspect of the present invention, the specific method for determining the generation of abnormal distribution signals is as follows:

[0025] Real-time data of cadmium in various forms in rice were acquired multiple times during the monitoring process, and the characteristic points of cadmium GHg for each monitored cadmium form were obtained. 实时 Where g represents different acquisition times, g = 1, 2, ..., y, y is a positive integer greater than 1, the spatial calibration points corresponding to the real-time monitoring link are obtained, and the GHg of each monitored cadmium morphology characteristic point is calculated. 实时 The distance Lg between each monitoring point and the corresponding spatial calibration point is used. If Lg is less than or equal to the preset radius β, the corresponding cadmium morphology feature point is determined to belong to the cadmium morphology distribution space corresponding to the real-time monitoring point. If Lg is greater than the preset radius β, the corresponding cadmium morphology feature point does not belong to the cadmium morphology distribution space corresponding to the real-time monitoring point. The ratio between the number of monitoring cadmium morphology feature points f belonging to the cadmium morphology distribution space corresponding to the real-time monitoring point and the number of acquisitions g is used as the location qualification rate JA corresponding to the real-time monitoring point. The same calculation and analysis method as the baseline topology coefficient T1 corresponding to the target point is used. The monitoring topology coefficient JB corresponding to the monitoring link is calculated. The absolute value of the difference JD between the benchmark topology coefficient and the monitoring topology coefficient corresponding to the monitoring link is obtained. The sum of the products of the reciprocal of the location qualification rate JA corresponding to the real-time monitoring link and the absolute value of the difference JD with the preset value coefficients ω1 and ω2 is calculated and used as the reasonable coefficient HL corresponding to the monitoring link. The specific values ​​of the preset value coefficients ω1 and ω2 are determined by relevant personnel according to actual needs, satisfying 1=ω1+ω2, ω1<ω2. When the reasonable coefficient HL is greater than the preset value Y1, an abnormal distribution signal is generated; otherwise, no processing is performed.

[0026] The IoT-based monitoring method for the intensive processing of rice includes the following steps:

[0027] Step 1: Obtain historical content data of cadmium in various forms at each stage of rice processing;

[0028] Step 2: Analyze the historical content data of cadmium in each form at each stage of rice refining process to obtain the spatial distribution of cadmium in each refining stage.

[0029] Step 3: Obtain the spatial calibration points corresponding to each finishing stage;

[0030] Step 4: Obtain the reference topology coefficients corresponding to each finishing stage;

[0031] Step 5: Analyze and acquire multiple cadmium morphology feature points of rice in the monitoring process to obtain the location pass rate of the monitoring process, acquire the monitoring topology coefficient, and determine the generation of abnormal distribution signals based on the location pass rate and monitoring topology coefficient of rice in the monitoring process.

[0032] Compared with the prior art, the present invention has the following beneficial effects:

[0033] This invention maps the content data of different cadmium forms in rice to a multi-dimensional feature space. Based on historical health data, it constructs a distribution space and spatial calibration points covering the distribution characteristics of cadmium forms, and further constructs a benchmark topological coefficient that can quantify the degree of distribution aggregation and uniformity. Finally, it calculates the location qualification rate and real-time monitoring topological coefficient through real-time data, and then weights and fuses them into a comprehensive reasonable coefficient. The threshold of this coefficient is used to determine the intelligent early warning of abnormal cadmium form distribution. This invention realizes real-time, multi-dimensional monitoring of cadmium forms during rice refining, and can quantitatively establish a health distribution benchmark based on topological features. On this basis, by comprehensively considering the spatial location conformity and topological structure deviation, it realizes intelligent judgment of abnormal cadmium form distribution and generates abnormal distribution signals. This invention achieves real-time monitoring of cadmium forms during rice refining, effectively improving the level of rice refining. Attached Figure Description

[0034] Figure 1 This is a schematic diagram of the system framework structure of the present invention;

[0035] Figure 2 This is a schematic diagram of the method framework structure of the present invention. Detailed Implementation

[0036] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0037] Example 1: Please refer to Figure 1This application provides an IoT-based monitoring system for the deep processing of rice, including:

[0038] The data acquisition module obtains historical content data of cadmium in various forms at each stage of rice processing.

[0039] It should be noted that each stage refers to the end of each stage, where sensors capable of providing near real-time total cadmium content or indicative heavy metal information are deployed, such as online X-ray fluorescence spectrometers or laser-induced breakdown spectrometers. These technologies can provide elemental content information, although they cannot directly distinguish the speciation, but can serve as a preliminary warning. For rapid speciation analysis, microfluidic chips, electrochemical sensor arrays, or electrochemical detectors / spectral detectors combined with rapid chromatographic separation are introduced to achieve minute-level cadmium speciation analysis. By combining the two, the monitoring of cadmium content in each speciation can be obtained. All of the above are existing and mature technologies, so they will not be elaborated on here.

[0040] The data acquisition module obtains historical cadmium content data at each stage of rice processing. Using sensor technologies such as online X-ray fluorescence spectrometry and laser-induced breakdown spectrometry, combined with microfluidic chips and electrochemical sensor arrays, near real-time monitoring of cadmium speciation is achieved.

[0041] The data analysis module analyzes the historical content data of various cadmium forms at each stage of rice processing to obtain the spatial distribution of cadmium forms at each stage. Specifically, the method is as follows:

[0042] Select one of the multiple finishing stages without replacement as the target stage;

[0043] Historical cadmium content data of various forms at the target processing stage were obtained from multiple batches of rice processing. These historical cadmium content data for each form at the target processing stage were then labeled Gi (GAi, GBi, GCI), where i represents different processing batches, and GBi, GBi, and GCI represent different forms of cadmium (e.g., GA represents Cd). 2+ GB stands for CdS, GC stands for Cd(OH)2;

[0044] The maximum values ​​GAmax, GBmax, and GCmax corresponding to different forms of cadmium GAi, GBi, and GCI in all batches were obtained, and Gi(GAi, GBi, GCI) was normalized by GHi(GAi / GAmax, GBi / GBmax, GBi / GCmax).

[0045] The cadmium morphology feature points GHi corresponding to each batch in the target stage are obtained. After discretization complex analysis of each cadmium morphology feature point GHi, the cadmium morphology distribution space Q1 corresponding to the target stage is generated. The specific method is as follows:

[0046] Centered on the cadmium morphology feature point GHi, draw spheres according to the preset radius β, and then obtain multiple spherical Ai corresponding to the covering complex Fi: Fi={(GHi,β)|i=1、2、……、n}, where i represents different processing batches and n represents the total number of batches;

[0047] The union of the geometric regions covered by each spherical Ai and the complex F is taken as the cadmium morphology distribution space Q1 corresponding to the target element.

[0048] The preset radius β is determined as follows:

[0049] The mean Gp of the distance between feature points of all batches and the standard deviation U of the distance between feature points of all batches are obtained according to GHi(GAi / GAmax, GBi / GBmax, GBi / GCmax). The sum of the mean Gp and twice the standard deviation is used as the preset radius β, i.e. β=Gp+2Gp.

[0050] Using the same analytical method, the cadmium speciation distribution space Qb corresponding to each refining stage can be obtained, where b represents different refining stages, b = 1, 2, ..., r, where r represents the total number of refining stages;

[0051] The spatial calibration point acquisition module analyzes and obtains the spatial calibration points corresponding to each refining stage of multiple batches of rice processing based on the cadmium morphology characteristic points at each refining stage. The specific method is as follows:

[0052] First, the mean values ​​of each cadmium morphology feature data GAi / GAmax, GBi / GBmax and GBi / GCmax are obtained from the cadmium morphology feature points GHi(GAi / GAmax, GBi / GBmax, GBi / GCmax) corresponding to each batch in the target stage, and then the spatial calibration point B1(BA, BB, BC) corresponding to the target stage is obtained.

[0053] Using the same analytical method, the spatial calibration point Bb corresponding to each finishing stage can be obtained;

[0054] The baseline topology coefficient acquisition module analyzes cadmium morphological feature points and spatial calibration points corresponding to each refining stage during multiple batches of rice processing to obtain the baseline topology coefficients for each refining stage. The specific method is as follows:

[0055] Obtain the cadmium morphology characteristic points GHi (GAi / GAmax, GBi / GBmax, GBi / GCmax) and the spatial calibration points B1 (BA, BB, BC) corresponding to each batch in the target stage.

[0056] The distance Di between each cadmium morphology feature point and the spatial calibration point is calculated using a distance formula. Based on a preset radius β, the cadmium morphology distribution space Q1 corresponding to the target stage is divided into a spherical neighborhood centered on the spatial calibration point B1. Cadmium morphology feature points with a distance Di less than or equal to the preset radius β are identified as feature points falling within the neighborhood of calibration point B1. The number e of feature points falling within the neighborhood of calibration point B1 is counted, and the number e is compared with the volume V of the neighborhood of calibration point B1. B1 The ratio between them is used as the neighborhood density P corresponding to the target element. B1 ; Neighborhood density P B1 It reflects the degree of cadmium aggregation near the calibration point;

[0057] Obtain the volume V of the neighborhood of calibration point B1 B1 The specific method is as follows:

[0058] pass Obtain the volume V of the neighborhood of calibration point B1 B1；

[0059] Through T1=α×P B1 +(1-α), calculate the baseline topology coefficient T1 corresponding to the target link, where α is a preset weight coefficient, and the value of α is 0.5 to 0.7;

[0060] The higher the baseline topology coefficient T1 value, the more highly concentrated and uniformly distributed the cadmium species is near the calibration point in the target process; the lower the baseline topology coefficient T1 value, the more dispersed and unevenly distributed the cadmium species is near the calibration point in the target process.

[0061] By using the same analysis method as that used to obtain the baseline topology coefficient T1 corresponding to the target stage, the baseline topology coefficient Tb corresponding to each finishing stage can be obtained.

[0062] The data analysis module performs normalization and discretization complex analysis on the collected data to generate the cadmium morphology distribution space and determine the spatial calibration points and reference topological coefficients.

[0063] The real-time monitoring module repeatedly acquires and analyzes real-time data on various forms of cadmium in rice during the monitoring process, obtaining multiple cadmium morphology feature points. It analyzes whether each feature point falls within the cadmium morphology distribution space of its corresponding monitoring stage, obtaining the location pass rate for that stage. Simultaneously, it acquires the monitoring topology coefficient based on these feature points. Based on the location pass rate and topology coefficient of rice within the monitoring stage, it determines whether the cadmium morphology distribution of rice is abnormal within the corresponding monitoring stage. An abnormal distribution signal is generated based on the determination result. The specific determination method is as follows:

[0064] Sensors deployed at various stages were used to repeatedly acquire real-time data on cadmium in different forms in rice during the monitoring process. The real-time cadmium content data were labeled as Gg. 实时 (GAg 实时 GBg 实时 GCg 实时 The cadmium morphology characteristic points GHg were obtained by normalizing the maximum values ​​GAmax, GBmax, and GCmax of cadmium morphology GAi, GBi, and GCI across all batches. 实时 (GAg 实时 / GAmax, GBg 实时 / GBmax,GCg 实时 / GCmax), where g represents the different number of acquisitions, g = 1, 2, ..., y, and y is a positive integer greater than 1;

[0065] Obtain the spatial calibration points corresponding to the real-time monitoring process, and calculate the GHg of each cadmium morphology characteristic point. 实时 (GAg 实时 / GAmax, GBg 实时 / GBmax,GCg 实时 The distance Lg between / GCmax and the spatial calibration point corresponding to the monitoring link is determined. When Lg is less than or equal to the preset radius β, it is determined that the corresponding cadmium morphology feature point belongs to the cadmium morphology distribution space corresponding to the real-time monitoring link. When Lg is greater than the preset radius β, the corresponding cadmium morphology feature point does not belong to the cadmium morphology distribution space corresponding to the real-time monitoring link. The ratio between the number of cadmium morphology feature points f belonging to the cadmium morphology distribution space corresponding to the real-time monitoring link and the number of acquisitions g is used as the location qualification rate JA corresponding to the real-time monitoring link.

[0066] Using the same calculation and analysis method as the baseline topology coefficient T1 corresponding to the target stage, the monitoring topology coefficient JB corresponding to the monitoring stage is calculated. The specific method is as follows:

[0067] To monitor cadmium morphological characteristic point GHg 实时Centered on a sphere with a preset radius θ, multiple GHg points are obtained to monitor cadmium morphology characteristics. 实时 A spherical Mg is centered, and multiple spherical Mg are obtained, each corresponding to a complex Vg: Vg={(Mg,θ)|g=1,2,……,y};

[0068] The union of the geometric regions covered by the complex Vg corresponding to each spherical Mg is taken as the cadmium speciation monitoring distribution space JC corresponding to the real-time monitoring link, i.e.

[0069] The preset radius θ is determined as follows;

[0070] All cadmium speciation characteristic points GHg were obtained based on GHi(GAi / GAmax, GBi / GBmax, GBi / GCmax). 实时 (GAg 实时 / GAmax, GBg 实时 / GBmax,GCg 实时 The standard deviation W1 and mean W2 of the distance between / GCmax) are used as the preset radius θ, i.e., θ = θ + 2W2.

[0071] The cadmium morphology monitoring distribution space JC corresponding to the real-time monitoring link is divided into a spherical neighborhood centered on the spatial calibration point corresponding to the monitoring link. The cadmium morphology feature points with a distance Lg less than or equal to the preset radius θ are determined as feature points falling into the neighborhood of the monitoring link. The number of cadmium morphology feature points falling into the neighborhood of the monitoring link z is counted. The ratio between the number z and the volume VG in the neighborhood of the monitoring link is taken as the monitoring density VM corresponding to the monitoring link.

[0072] The monitoring topology coefficient JB corresponding to the monitoring link is calculated using JB = α × VM + (1 - α). ；

[0073] The specific method for obtaining the volume VG within the neighborhood of the monitoring link is as follows:

[0074] pass

[0075] Obtain the absolute value JD of the difference between the baseline topology coefficient and the monitoring topology coefficient corresponding to the monitoring link. Calculate the sum of the products of the reciprocal of the location qualification rate JA corresponding to the real-time monitoring link and the absolute value JD of the difference with the preset value coefficients ω1 and ω2 respectively, and use it as the reasonable coefficient HL corresponding to the monitoring link. The specific values ​​of the preset value coefficients ω1 and ω2 are determined by relevant personnel according to actual needs, satisfying 1=ω1+ω2, ω1<ω2.

[0076] When the reasonable coefficient HL is greater than the preset value Y1, an abnormal distribution signal is generated; otherwise, no processing is performed. The specific value of the preset value Y1 is determined by relevant personnel according to actual needs.

[0077] It should be noted that if the location qualification rate JA is 0, then 1 / JA will be infinite, which will directly cause HL to become infinite, thus inevitably triggering an abnormal signal. When JA is 0, an abnormal distribution signal will be generated directly.

[0078] The real-time monitoring module acquires real-time data multiple times, analyzes whether the monitored cadmium morphology feature points belong to the corresponding cadmium morphology distribution space, calculates the location qualification rate and monitoring topology coefficient, thereby determining whether the cadmium morphology distribution is abnormal and generating an abnormal distribution signal.

[0079] This technology enables real-time monitoring and precise analysis of cadmium speciation during rice processing, allowing for timely detection of abnormal cadmium content, preventing batch contamination, reducing resource waste, further ensuring the stability and safety of the processing, and improving the quality of rice products.

[0080] Example 2: As an example 2 of the present invention, please refer to... Figure 2 A method for monitoring the deep processing of rice based on the Internet of Things (IoT) is provided. This method is used to implement the aforementioned disclosed IoT-based monitoring system for the deep processing of rice, specifically including:

[0081] Step 1: Obtain historical content data of cadmium in various forms at each stage of rice processing;

[0082] Step Two: Analyze the historical content data of cadmium in various forms at each stage of rice processing to obtain the spatial distribution of cadmium forms in each stage. The specific method is as follows:

[0083] Select one of the multiple finishing stages without replacement as the target stage;

[0084] The maximum values ​​GAmax, GBmax, and GCmax corresponding to different forms of cadmium GAi, GBi, and GCI in all batches were obtained, and Gi(GAi, GBi, GCI) was normalized by GHi(GAi / GAmax, GBi / GBmax, GBi / GCmax).

[0085] Centered on the cadmium morphology feature point GHi, spheres are drawn according to a preset radius β, thereby obtaining multiple spherical Ai corresponding to the covering complex F. The union of the geometric regions covered by each spherical Ai corresponding to the covering complex F is taken as the cadmium morphology distribution space Q1 corresponding to the target segment.

[0086] Using the same analytical method, the cadmium speciation distribution space Qb corresponding to each refining stage can be obtained, where b represents different refining stages, b = 1, 2, ..., r, where r represents the total number of refining stages;

[0087] The spatial calibration point acquisition module analyzes and obtains the spatial calibration points corresponding to each refining stage of multiple batches of rice processing based on the cadmium morphology characteristic points at each refining stage. The specific method is as follows:

[0088] First, the mean values ​​of each cadmium morphology characteristic data GAi / GAmax, GBi / GBmax and GBi / GCmax are obtained from the cadmium morphology characteristic points GHi (GAi / GAmax, GBi / GBmax, GBi / GCmax) corresponding to each batch in the target stage. Then, the spatial calibration point B1 (BA, BB, BC) corresponding to the target stage is obtained. By adopting the same analysis method, the spatial calibration point Bb corresponding to each finishing stage can be obtained.

[0089] The baseline topology coefficient acquisition module analyzes cadmium morphological feature points and spatial calibration points corresponding to each refining stage during multiple batches of rice processing to obtain the baseline topology coefficients for each refining stage. The specific method is as follows:

[0090] Obtain the cadmium morphology characteristic points GHi (GAi / GAmax, GBi / GBmax, GBi / GCmax) and the spatial calibration points B1 (BA, BB, BC) corresponding to each batch in the target stage.

[0091] The distance Di between each cadmium morphology feature point and the spatial calibration point is calculated using a distance formula. Based on a preset radius β, the cadmium morphology distribution space Q1 corresponding to the target stage is divided into a spherical neighborhood centered on the spatial calibration point B1. Cadmium morphology feature points with a distance Di less than or equal to the preset radius β are identified as feature points falling within the neighborhood of calibration point B1. The number e of feature points falling within the neighborhood of calibration point B1 is counted, and the number e is compared with the volume V of the neighborhood of calibration point B1. B1 The ratio between them is used as the neighborhood density P corresponding to the target element. B1 ; Neighborhood density P B1 It reflects the degree of cadmium aggregation near the calibration point;

[0092] Through T1=α×P B1 +(1-α), calculate the reference topology coefficient T1 corresponding to the target stage; using the same analysis method as for obtaining the reference topology coefficient T1 corresponding to the target stage, the reference topology coefficient Tb corresponding to each finishing stage can be obtained;

[0093] The real-time monitoring module repeatedly acquires and analyzes real-time data on various forms of cadmium in rice during the monitoring process, obtaining multiple cadmium morphology feature points. It analyzes whether each feature point falls within the cadmium morphology distribution space of its corresponding monitoring stage, obtaining the location pass rate for that stage. Simultaneously, it acquires the monitoring topology coefficient based on these feature points. Based on the location pass rate and topology coefficient of rice within the monitoring stage, it determines whether the cadmium morphology distribution of rice is abnormal within the corresponding monitoring stage. An abnormal distribution signal is generated based on the determination result. The specific determination method is as follows:

[0094] Obtain the spatial calibration points corresponding to the real-time monitoring process, and calculate the GHg of each cadmium morphology characteristic point. 实时 The distance Lg between each monitoring point and the corresponding spatial calibration point is used. If Lg is less than or equal to the preset radius β, the corresponding cadmium morphology feature point is determined to belong to the cadmium morphology distribution space corresponding to the real-time monitoring point. If Lg is greater than the preset radius β, the corresponding cadmium morphology feature point does not belong to the cadmium morphology distribution space corresponding to the real-time monitoring point. The ratio between the number of monitoring cadmium morphology feature points f belonging to the cadmium morphology distribution space corresponding to the real-time monitoring point and the number of acquisitions g is used as the location qualification rate JA corresponding to the real-time monitoring point. The monitoring topology coefficient JB corresponding to the monitoring point is calculated using the same calculation and analysis method as the baseline topology coefficient T1 corresponding to the target point. The specific method is as follows:

[0095] Obtain the absolute value JD of the difference between the baseline topology coefficient and the monitoring topology coefficient corresponding to the monitoring link. Calculate the sum of the products of the reciprocal of the location qualification rate JA corresponding to the real-time monitoring link and the absolute value JD of the difference with the preset value coefficients ω1 and ω2 respectively, and use it as the reasonable coefficient HL corresponding to the monitoring link. The specific values ​​of the preset value coefficients ω1 and ω2 are determined by relevant personnel according to actual needs, satisfying 1=ω1+ω2, ω1<ω2.

[0096] When the reasonable coefficient HL is greater than the preset value Y1, an abnormal distribution signal is generated; otherwise, no processing is performed. The specific value of the preset value Y1 is determined by relevant personnel according to actual needs.

[0097] It should be noted that if the location qualification rate JA is 0, then 1 / JA will be infinite, which will directly cause HL to become infinite, thus inevitably triggering an abnormal signal. When JA is 0, an abnormal distribution signal will be generated directly.

[0098] The data on the content of different forms of cadmium in rice are mapped to a multidimensional feature space. Based on historical health data, a "distribution space" and "spatial calibration points" covering the distribution characteristics of cadmium forms are constructed. Furthermore, a "benchmark topology coefficient" that can quantify the degree of distribution aggregation and uniformity is constructed. Finally, the location qualification rate and real-time monitoring topology coefficient are calculated through real-time data, and they are weighted and integrated into a comprehensive "reasonable coefficient". The threshold of this coefficient is used to determine the intelligent early warning of abnormal cadmium form distribution.

[0099] This technology enables real-time, multi-dimensional monitoring of cadmium speciation during rice refining. It can quantitatively establish a health distribution benchmark based on topological features. Furthermore, by comprehensively considering spatial location conformity and topological deviation, it can intelligently determine abnormal cadmium speciation distribution, thereby enabling enterprises to proactively intervene in processes and control risks, effectively improving the level of rice refining, and providing data support for intelligent management and continuous optimization of the production process.

[0100] Example 3: As Example 3 of the present invention, in specific implementation, compared with Example 1 and Example 2, the technical solution of this example is to combine the solutions of Example 1 and Example 2.

[0101] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters and thresholds in the formulas are set by those skilled in the art according to the actual situation.

[0102] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A monitoring system for the intensive processing of rice based on the Internet of Things, characterized in that, include; The data acquisition module obtains historical content data of cadmium in various forms at each stage of rice processing. The data analysis module selects one of the multiple finishing processes without replacement as the target process. Historical cadmium content data of various forms at the target stage in multiple batches of rice processing were obtained. Historical cadmium content data Gi (GAi, GBi, GCI) of various forms of cadmium at the target stage were obtained for each batch of rice during processing. After normalization, cadmium morphology feature points GHi corresponding to each batch at the target stage were obtained. After discretization complex analysis of each cadmium morphology feature point GHi, the cadmium morphology distribution space Q1 corresponding to the target stage was generated, where i refers to different processing batches, i=1,2,...,n, n refers to the total number of batches, and GAi, GBi and GCI refer to different forms of cadmium. Using the same analytical method, the cadmium speciation distribution space Qb corresponding to each refining stage can be obtained, where b represents different refining stages, b=1, 2, ..., r, where r represents the total number of refining stages; The spatial calibration point acquisition module first obtains the mean values ​​of each cadmium morphology feature data GAi / GAmax, GBi / GBmax, and GBi / GCmax in the cadmium morphology feature points GHi (GAi / GAmax, GBi / GBmax, GBi / GCmax) corresponding to each batch in the target stage, and then obtains the spatial calibration point B1 (BA, BB, BC) corresponding to the target stage. Using the same analysis method, the spatial calibration point Bb corresponding to each finishing stage can be obtained. The baseline topology coefficient acquisition module obtains the cadmium morphology feature points GHi (GAi / GAmax, GBi / GBmax, GBi / GCmax) and spatial calibration points B1 (BA, BB, BC) corresponding to each batch in the target stage. The distance Di between each cadmium morphology feature point and the spatial calibration point is calculated using a distance formula. Based on a preset radius β, the cadmium morphology distribution space Q1 corresponding to the target stage is divided into a spherical neighborhood centered on the spatial calibration point B1. Cadmium morphology feature points with a distance Di less than or equal to the preset radius β are considered to fall within the neighborhood of calibration point B1. The number e of feature points falling within the neighborhood of calibration point B1 is counted, and the number e is compared with the volume V of the neighborhood of calibration point B1. B1 The ratio between them is used as the neighborhood density P corresponding to the target element. B1 ; through T1=α×P B1 + (1-α) to calculate the baseline topology coefficient T1 corresponding to the target stage, where α ranges from 0.5 to 0.7; using the same analysis method as for obtaining the baseline topology coefficient T1 corresponding to the target stage, the baseline topology coefficient Tb corresponding to each finishing stage can be obtained; The real-time monitoring module repeatedly acquires real-time data on various forms of cadmium in rice during the monitoring process, obtaining GHg characteristic points for each monitored cadmium form. 实时 Where g represents different acquisition times, g = 1, 2, ..., y, y is a positive integer greater than 1, obtaining the spatial calibration points corresponding to the real-time monitoring link, and calculating GHg for each monitored cadmium morphology characteristic point. 实时 The distance Lg between each monitoring point and the corresponding spatial calibration point is used. If Lg is less than or equal to the preset radius β, the corresponding cadmium morphology feature point is determined to belong to the cadmium morphology distribution space corresponding to the real-time monitoring point. If Lg is greater than the preset radius β, the corresponding cadmium morphology feature point does not belong to the cadmium morphology distribution space corresponding to the real-time monitoring point. The ratio between the number of monitoring cadmium morphology feature points f belonging to the cadmium morphology distribution space corresponding to the real-time monitoring point and the number of acquisitions g is used as the location qualification rate JA corresponding to the real-time monitoring point. The same calculation and analysis method as the baseline topology coefficient T1 corresponding to the target point is used. The monitoring topology coefficient JB corresponding to the monitoring link is calculated. The absolute value of the difference JD between the benchmark topology coefficient and the monitoring topology coefficient corresponding to the monitoring link is obtained. The sum of the products of the reciprocal of the location qualification rate JA corresponding to the real-time monitoring link and the absolute value of the difference JD with the preset value coefficients ω1 and ω2 is calculated and used as the reasonable coefficient HL corresponding to the monitoring link. The specific values ​​of the preset value coefficients ω1 and ω2 are determined by relevant personnel according to actual needs, satisfying 1=ω1+ω2, ω1<ω2. When the reasonable coefficient HL is greater than the preset value Y1, an abnormal distribution signal is generated; otherwise, no processing is performed.

2. The IoT-based rice deep processing monitoring system according to claim 1, characterized in that, The specific method of normalization is as follows: The maximum values ​​GAmax, GBmax, and GCmax corresponding to different forms of cadmium GAi, GBi, and GCI in all batches were obtained, and Gi(GAi, GBi, GCI) was normalized by GHi(GAi / GAmax, GBi / GBmax, GBi / GCmax).

3. The IoT-based rice deep processing monitoring system according to claim 1, characterized in that, The specific method for discretizing and performing complex analysis on each cadmium morphological feature point GHi is as follows: Centered on the cadmium morphology feature point GHi, a sphere is drawn according to the preset radius β, and then multiple spheres Ai are obtained, each corresponding to a complex Fi. The union of the geometric regions covered by the complex F corresponding to each sphere Ai is taken as the cadmium morphology distribution space Q1 corresponding to the target segment.

4. The IoT-based rice deep processing monitoring system according to claim 3, characterized in that, The preset radius β is determined as follows; The mean Gp of the distance between all cadmium morphological feature points and the standard deviation of the distance between all batch feature points are obtained based on GHi (GAi / GAmax, GBi / GBmax, GBi / GCmax). The sum of the mean Gp and twice the standard deviation is used as the preset radius β, i.e., β = Gp + 2Gp.

5. The IoT-based rice deep processing monitoring system according to claim 4, characterized in that, Obtain the volume V of the neighborhood of calibration point B1 B1 The specific method is as follows: pass Obtain the volume V of the neighborhood of calibration point B1. B1 .

6. A monitoring method for the intensive processing of rice based on the Internet of Things, characterized in that, This method implements the IoT-based rice deep processing monitoring system according to any one of claims 1-5, comprising: Step 1: Obtain historical content data of cadmium in various forms at each stage of rice processing; Step 2: Analyze the historical content data of cadmium in each form at each stage of rice refining process to obtain the spatial distribution of cadmium in each refining stage. Step 3: Obtain the spatial calibration points corresponding to each finishing stage; Step 4: Obtain the reference topology coefficients corresponding to each finishing stage; Step 5: Analyze and acquire multiple cadmium morphology feature points of rice in the monitoring process to obtain the location pass rate of the monitoring process, acquire the monitoring topology coefficient, and determine the generation of abnormal distribution signals based on the location pass rate and monitoring topology coefficient of rice in the monitoring process.

Images