An intelligent safety risk monitoring and early warning management system for the grain industry

Abstract

The application discloses a kind of intelligent safety risk monitoring and early warning management systems for grain industry, it is related to the technical field of intelligent management of grain storage, comprising: deploying a plurality of physical state response units in grain pile, based on non-contact scanning device, a plurality of physical state response units are scanned in global, to obtain the current physical state information of each physical state response unit;Based on the current physical state information, determine the state offset of each physical state response unit;Based on the state offset, determine risk determination result.Non-contact scanning mode does not need to be arranged in grain pile Any cable or communication module, avoids signal attenuation and transmission delay problem, realizes the periodicity, continuity monitoring to grain pile state.

Description

Technical Field

[0001] This invention relates to the field of intelligent management technology for grain storage, specifically to an intelligent safety risk monitoring and early warning management system for the grain industry. Background Technology

[0002] During grain storage, the grain pile is prone to localized mold, heat generation, and pest infestation due to factors such as microbial activity, pest growth, and grain respiration. If these risks are not detected and addressed in a timely manner, they may lead to large-scale grain spoilage and significant economic losses. Therefore, real-time monitoring and early warning of the risk status inside the grain pile are of great importance.

[0003] Existing methods mainly employ the placement of electronic sensors inside the grain pile or on the silo walls. Common approaches include: placing temperature and humidity sensors to monitor changes in the temperature and humidity fields of the grain pile; placing gas sensors to monitor the concentration of volatile gases such as carbon dioxide and ammonia; and placing pressure sensors to monitor the pressure distribution of the grain pile.

[0004] However, the electronic sensors in the existing methods require continuous power supply, and the power supply lines are complicated to lay out in the closed environment of the grain pile. Furthermore, the sensors are prone to corrosion and failure due to long-term exposure to high humidity and dust, resulting in high maintenance costs. The data transmission of a large number of sensors relies on wired networks or wireless communication modules, and the signal attenuation is severe inside the grain pile, making it difficult to guarantee the stability of data transmission. Summary of the Invention

[0005] (a) Technical problems to be solved To address the shortcomings of existing technologies, this invention provides an intelligent safety risk monitoring and early warning management system for the grain industry.

[0006] (II) Technical Solution To achieve the above objectives, the present invention provides the following technical solution: an intelligent safety risk monitoring and early warning management system for the grain industry, specifically comprising: Sensor deployment module: Several physical state response units are deployed within the grain pile, and the physical state response units are composed of physically variable materials that are sensitive to risk-induced mutation factors; Data acquisition module: Based on a non-contact scanning device, it performs a full-domain scan of several physical state response units to obtain the current physical state information of each physical state response unit; State offset calculation module: Based on the current physical state information, determine the state offset of each physical state response unit. The state offset is used to characterize the degree of irreversible change of its physical state compared to the initial state. Risk analysis module: Based on the state offset, determine the risk assessment result, which includes the risk type, risk level, and risk location.

[0007] Preferably, the sensor deployment module deploys physical state response units in the form of three-dimensional mesh nodes inside the grain pile; Based on the deployment strategy, a unique spatial code is generated for each physical state response unit, and the unique spatial code is associated with its three-dimensional coordinate information in the grain pile; Based on unique spatial coding, an initial state database is established. The initial state database is used to record the initial physical state parameters of each physical state response unit during deployment.

[0008] Preferably, the data acquisition module acquires a broadband excitation beam emitted by a non-contact scanning device. The broadband excitation beam sequentially covers the spatial location of each physical state response unit based on a preset scanning path, and the spot diameter of the broadband excitation beam matches the size of the physical state response unit. Based on a broadband excitation beam, optical excitation is performed on the surface of each physical state response unit to obtain the feedback signal generated by each physical state response unit. The feedback signal is received sequentially by the optical receiver built into the non-contact scanning device according to the order of the preset scanning path, and each feedback signal is associated with and stored with the current scanning position information. Based on the feedback signal, the current optical characteristic parameter set of each physical state response unit is determined. The current optical characteristic parameter set is obtained by analyzing the feedback signal. The analysis process includes extracting at least one of the following from the feedback signal: the reflection spectrum intensity distribution sequence, the fluorescence peak wavelength value, the fluorescence intensity attenuation value, and the surface morphology interference fringe spacing value. The analyzed data is then associated and stored with the unique spatial code of the physical state response unit.

[0009] Preferably, the state offset calculation module extracts the initial optical feature parameter set corresponding to the physical state response unit from the initial state database based on a unique spatial code. The initial optical feature parameter set includes the initial reflectance spectral intensity distribution sequence, the initial fluorescence peak wavelength value, the initial fluorescence intensity value, the initial interference fringe spacing value, and the initial interference fringe contrast value. Based on the current optical feature parameter set and the initial optical feature parameter set, a parameter-by-parameter comparison is performed to generate feature difference values ​​in at least one dimension. The parameter-by-parameter comparison includes: comparing the current reflectance spectral intensity distribution sequence with the initial reflectance spectral intensity distribution sequence to form a reflectance spectral difference sequence; comparing the current fluorescence peak wavelength value with the initial fluorescence peak wavelength value to generate a fluorescence peak wavelength shift; comparing the current fluorescence intensity value with the initial fluorescence intensity value to generate a fluorescence intensity attenuation value; comparing the current interference fringe spacing value with the initial interference fringe spacing value to generate an interference fringe spacing change; and comparing the current interference fringe contrast with the initial interference fringe contrast to generate an interference fringe contrast change. Based on the feature difference value of at least one dimension, the comprehensive state offset of the physical state response unit is determined. The process of determining the comprehensive state offset includes normalizing the feature difference value of each dimension to obtain the normalized change degree, multiplying the normalized change degree of each dimension by a preset weight coefficient to obtain the weighted change degree, and summing the weighted change degrees of all dimensions to obtain the comprehensive state offset.

[0010] Preferably, the risk analysis module obtains the comprehensive state offset of multiple physical state response units and their corresponding unique spatial codes, and maps the comprehensive state offset of each physical state response unit to the three-dimensional coordinate position indicated by the unique spatial code to form a risk offset spatial distribution map. Based on the risk offset spatial distribution map, spatial clustering analysis is performed. The comprehensive state offset is compared with the first preset threshold. Spatial coordinate points with a comprehensive state offset greater than or equal to the first preset threshold are marked as risk points, and spatial coordinate points with a comprehensive state offset less than the first preset threshold are marked as non-risk points. Through spatial connectivity judgment, interconnected risk points are merged into continuous regions, and the continuous regions formed by physical state response units with comprehensive state offsets exceeding the first preset threshold are determined. Based on a continuous region, the length, width, and height coordinates of all risk points within the region are obtained. The length, width, and height coordinates are summed to obtain the total length coordinates, width coordinates, and height coordinates, respectively. The total number of risk points is counted. The average length coordinate is obtained by dividing the total length coordinate by the total number of risk points. The average width coordinate is obtained by dividing the total width coordinate by the total number of risk points. The average height coordinate is obtained by dividing the total height coordinate by the total number of risk points. The three-dimensional coordinate point formed by the average length coordinate, average width coordinate, and average height coordinate is determined as the geometric center of the continuous region. The geometric center is used as the risk location. Based on a continuous region, the comprehensive state offset values ​​of risk points within the continuous region are obtained. The minimum, maximum, and average values ​​of the comprehensive state offset are determined. The average or maximum value is compared with a preset risk level classification rule to determine the numerical range to which the average or maximum value belongs. The risk level corresponding to the numerical range is determined as the risk level of the continuous region. The feature difference values ​​of at least one dimension of at least one physical state response unit within the continuous region are obtained. The feature difference values ​​of at least one dimension include at least one of the following: reflectance spectrum difference sequence, fluorescence peak wavelength offset, fluorescence intensity attenuation value, interference fringe spacing change, and interference fringe contrast change. Based on feature difference values ​​in at least one dimension, feature difference value combination types are formed by combining them in a preset order. The feature difference value combination types are compared with the pre-defined correspondence between feature difference value combination types and risk types. The risk types that match the feature difference value combination types are determined as the risk types of continuous regions.

[0011] (III) Beneficial Effects This invention provides an intelligent safety risk monitoring and early warning management system for the grain industry, which has the following beneficial effects: This invention achieves passive risk monitoring inside grain piles by deploying physical state response units made of materials with variable physical states. The physical state response units require no power supply or data transmission, reducing deployment and maintenance costs. The current physical state information of each physical state response unit is acquired through a non-contact optical scanning device, enabling full-area, real-time acquisition of risk states inside the grain pile. The non-contact scanning method eliminates the need for any cables or communication modules inside the grain pile, avoiding signal attenuation and transmission delay issues, and enabling periodic and continuous monitoring of the grain pile state. By associating the state offset of the physical state response units with their spatial codes, a risk offset spatial distribution map is formed. Combined with spatial clustering analysis, continuous regions are determined, enabling precise location of risk positions. By determining the geometric center of the continuous regions, the three-dimensional coordinate position of risk events inside the grain pile can be accurately identified. Attached Figure Description

[0012] Figure 1 This is a system block diagram of the present invention. Detailed Implementation

[0013] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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.

[0014] Please see Figure 1 This invention provides an intelligent safety risk monitoring and early warning management system for the grain industry, comprising: Sensor deployment module: Several physical state response units are deployed within the grain pile, and the physical state response units are composed of physically variable materials that are sensitive to risk-induced mutation factors; In this embodiment of the invention, the sensor deployment module is specifically described. The sensor deployment module deploys physical state response units in the form of three-dimensional mesh nodes inside the grain pile. Based on the deployment strategy, a unique spatial code is generated for each physical state response unit, and the unique spatial code is associated with its three-dimensional coordinate information in the grain pile; Based on unique spatial coding, an initial state database is established. The initial state database is used to record the initial physical state parameters of each physical state response unit during deployment.

[0015] It should be noted that, based on the geometric dimensions of the grain pile and the monitoring resolution requirements, the grain pile space is discretized into a three-dimensional grid coordinate system, which consists of grid nodes arranged at equal intervals along the length, width, and height of the grain pile. Each grid node has a unique coordinate value in the three-dimensional coordinate system, corresponding to its length, width and height coordinate values ​​respectively. The deployment strategy stipulates that a physical state response unit is buried at the location of each grid node, so that the physical state response units are distributed in a three-dimensional spatial array inside the grain pile. Based on the deployment strategy, a unique spatial code is generated for each physical state response unit. The unique spatial code is associated with its three-dimensional coordinate information in the grain pile. The unique spatial code is achieved by setting a micro-physical structure array with a specific arrangement on the surface of each physical state response unit. The micro-physical structure array is composed of multiple micron-level protrusions, pits or reflective surfaces arranged according to a preset binary coding rule. During the generation process, the three-dimensional coordinate values ​​of the grid nodes corresponding to the physical state response unit are converted into a unique encoding sequence, and then the encoding sequence is physically solidified on the surface of the physical state response unit in the form of a micro physical structure array. The micro-physical structure array carried on the surface of each physical state response unit constitutes its unique spatial identifier. When the micro-physical structure array is read by a non-contact scanning device, the three-dimensional coordinate information of the physical state response unit can be decoded. Based on a unique spatial code, an initial state database is established. The initial state database is used to record the initial physical state parameters of each physical state response unit during deployment. After the deployment of the physical state response units is completed, an initial full-domain scan of all physical state response units is performed using a non-contact scanning device. The initial full-domain scanning process includes: a non-contact scanning device emitting a broadband excitation beam to sequentially irradiate the surface of each physical state response unit. Under the irradiation of the broadband excitation beam, each physical state response unit generates a corresponding feedback signal based on the initial physical state of its surface micro-physical structure array and sensitive material. The non-contact scanning device receives the feedback signal and analyzes it to obtain the initial optical characteristic parameter set of each physical state response unit. The initial optical characteristic parameter set includes at least the reflection spectrum intensity value, the fluorescence spectrum peak wavelength value and / or the surface morphology interference fringe spacing value. Using the unique spatial code of each physical state response unit as an index identifier, the initial optical feature parameter set of the physical state response unit is associated and stored with its unique spatial code to form an initial state database. The initial state database records the initial physical state reference value of all physical state response units at the time of deployment, which serves as the basis for comparison in subsequent judgment of state offset.

[0016] The enhanced response unit is composed of a material with a variable physical state that is sensitive to risk-induced mutation factors, specifically including: Moisture / humidity sensitive physical state variable materials: Physical state variable materials include hygroscopic expansion type materials, specifically cross-linked polyacrylate polymer materials or modified cellulose materials. When these materials come into contact with moisture accumulation in the local grain pile due to microbial metabolic activities or grain respiration, irreversible volume expansion occurs. The volume expansion causes deformation of the micro-physical structure array set on the surface of the physical state response unit, specifically manifested as an increase in the height of micro-protrusions or a change in the depth of pits, thereby changing its reflection characteristics to a broadband excitation beam, forming a shift in the reflection spectrum or a change in the spacing of the surface morphology interference fringes; Variable physical state materials sensitive to volatile organic compounds: The variable physical state materials include colorimetric reaction materials, specifically porous carrier materials or metal-organic framework materials loaded with acid-base indicators. When these materials come into contact with specific volatile organic compounds (such as hexanal, octenal, carbon dioxide or ammonia) produced locally by mold or pest metabolism in grain piles, an irreversible chemical colorimetric reaction occurs. The colorimetric reaction leads to a change in the surface color of the physical state response unit, specifically manifested as a shift in the characteristic absorption peak of the reflectance spectrum in the visible or near-infrared band, thereby changing its reflectance characteristics to a broadband excitation beam, forming a reflectance spectrum shift. Temperature / localized heating-sensitive material with variable physical state: The material with variable physical state includes thermo-induced phase change material, specifically low-melting-point alloy material or thermochromic material. When these materials come into contact with abnormal heat generated by the localized proliferation of microorganisms or pest activity in the grain pile, they undergo irreversible phase change or color change. Low-melting-point alloy material melts and then solidifies when it reaches its melting point, resulting in a permanent change in the geometry of the micro-physical structure array. Thermochromic material undergoes molecular structure rearrangement when it reaches its color change temperature, resulting in an irreversible shift of the characteristic peaks of the reflection spectrum. The above materials are combined according to a preset hierarchical structure to form a physical state response unit, specifically including: The first layer is a selectively permeable membrane layer, made of polydimethylsiloxane or polytetrafluoroethylene, which allows volatile organic compounds and water molecules of a specific molecular weight to pass through while blocking grain particles. The second layer is a sensitive medium layer, which is disposed inside the first layer and is composed of at least one of the aforementioned hygroscopic expansion type material, color reaction type material and / or thermally induced phase change type material; The third layer is an optical identification layer, which is set on the surface of the second layer and is composed of a micro-nano structure array. It is used to synchronously change the optical reflection properties of the second layer when the shape or color changes.

[0017] Data acquisition module: Based on a non-contact scanning device, it performs a full-domain scan of several physical state response units to obtain the current physical state information of each physical state response unit; In this embodiment of the invention, the data acquisition module needs to be specifically described. The data acquisition module acquires a broadband excitation beam emitted by a non-contact scanning device. The broadband excitation beam sequentially covers the spatial position of each physical state response unit based on a preset scanning path, and the spot diameter of the broadband excitation beam matches the size of the physical state response unit. Based on a broadband excitation beam, optical excitation is performed on the surface of each physical state response unit to obtain the feedback signal generated by each physical state response unit. The feedback signal is received sequentially by the optical receiver built into the non-contact scanning device according to the order of the preset scanning path, and each feedback signal is associated with and stored with the current scanning position information. Based on the feedback signal, the current optical characteristic parameter set of each physical state response unit is determined. The current optical characteristic parameter set is obtained by analyzing the feedback signal. The analysis process includes extracting at least one of the following from the feedback signal: the reflection spectrum intensity distribution sequence, the fluorescence peak wavelength value, the fluorescence intensity attenuation value, and the surface morphology interference fringe spacing value. The analyzed data is then associated and stored with the unique spatial code of the physical state response unit.

[0018] It should be noted that the non-contact scanning device is set at a preset position above the grain pile or inside the grain silo. It integrates a broadband light source, which can emit a continuous spectral beam covering the ultraviolet, visible and near-infrared light bands. When scanning is started, the non-contact scanning device moves according to a preset scanning path. The preset scanning path is set based on the three-dimensional grid node positions determined in the aforementioned deployment strategy, ensuring that the broadband excitation beam can sequentially cover the spatial position of each physical state response unit. The emission angle and spot diameter of the broadband excitation beam are pre-calibrated so that when the beam irradiates the surface of the grain pile, it forms an irradiation spot that matches the size of the physical state response unit, avoiding overlap or omission of the beam irradiation range. Based on a broadband excitation beam, optical excitation is performed on the surface of each physical state response unit to obtain the feedback signal generated by each physical state response unit. When the broadband excitation beam irradiates the surface of the physical state response unit, the beam interacts with the current physical state of the micro physical structure array set on the unit surface and the internal sensitive medium layer. For micro-physical structure arrays, broadband excitation beams are reflected and interfered on their surface, generating reflected beams carrying structural morphology information. For sensitive medium layers, if they have undergone irreversible physical state changes (such as volume expansion or color change) due to contact risk mutation factors, broadband excitation beams will generate reflection, fluorescence, or absorption spectra on the surface of the medium layer corresponding to the changed state. The optical receiver built into the non-contact scanning device synchronously receives the aforementioned reflected beam, fluorescent beam, and / or scattered beam, converts these optical signals into electrical signals, and forms a feedback signal for each physical state response unit. During the scanning process, the optical receiver receives the feedback signal of each physical state response unit in sequence according to the order of the preset scanning path, and stores the feedback signal in association with the current scanning position information to ensure that the feedback signal corresponds to the unique spatial code of the physical state response unit. The feedback signal corresponding to each physical state response unit is analyzed and processed. The analysis process includes: after obtaining the feedback signal corresponding to each physical state response unit, firstly, band separation processing is performed. The feedback signal is collected by the optical receiver of the non-contact scanning device. The feedback signal contains reflection spectrum components, fluorescence spectrum components, and interference fringe components. The feedback signal is separated according to the preset band division rules. Specifically, the signal components with wavelength ranges between 200 nm and 400 nm are classified as ultraviolet reflection spectrum components, the signal components with wavelength ranges between 400 nm and 780 nm are classified as visible reflection spectrum components and fluorescence spectrum components. Among them, the signal components continuously collected after the excitation beam stops are classified as fluorescence spectrum components by time resolution technology, the signal components with wavelength ranges between 780 nm and 2500 nm are classified as near-infrared reflection spectrum components, and the periodic intensity change components generated by the micro-physical structure array in the feedback signal are classified as interference fringe components. Through band separation, the composite feedback signal is decomposed into multiple single-type signal components. Based on the separated ultraviolet reflectance spectral components, visible reflectance spectral components, and near-infrared reflectance spectral components, reflectance spectral intensity distribution data is extracted. The specific extraction steps are as follows: the ultraviolet reflectance spectral components, visible reflectance spectral components, and near-infrared reflectance spectral components are merged and sorted according to wavelength values ​​to form a continuous sequence from the starting wavelength to the ending wavelength. The continuous sequence covers the complete spectral range from 200 nm to 2500 nm. At each wavelength position in the continuous sequence, the light intensity value corresponding to the wavelength is measured. The light intensity value is the reflected light energy received by the optical receiver of the non-contact scanning device at that wavelength. The reference intensity value of the incident broadband excitation beam at that wavelength is obtained. This reference intensity value is obtained through pre-calibration, that is, the initial emission intensity of the broadband excitation beam at each wavelength is measured without any physical state response unit blocking it. The ratio of the light intensity value at a given wavelength to the reference intensity value of the incident broadband excitation beam at that wavelength is used as the reflectivity value at that wavelength. This reflectivity value is dimensionless and ranges from 0 to 1, characterizing the reflectivity capability of the physical state response unit surface to light of that wavelength. All wavelength positions and their corresponding reflectivity values ​​are arranged in ascending order of wavelength to form a reflectivity spectrum intensity distribution sequence. The reflectivity spectrum intensity distribution sequence records the reflectivity characteristics of the physical state response unit surface to light of different wavelengths in the current state. Based on the separated fluorescence spectral components, fluorescence spectral feature parameters are extracted. The specific extraction steps are as follows: background noise is filtered out from the fluorescence spectral components. A preset noise threshold is used for judgment. Signal components with signal intensity lower than the threshold are identified as non-fluorescent signals introduced by ambient light or system noise and are removed. Effective fluorescence signals with signal intensity higher than the threshold are retained. In the effective fluorescence signal, the peak position is identified by traversing the signal intensity values ​​at each wavelength position, determining the wavelength value corresponding to the maximum signal intensity, and defining this wavelength value as the fluorescence peak wavelength value. This parameter characterizes the emission characteristics of the fluorescent substance. The signal intensity value at the peak wavelength position is measured and determined as the fluorescence intensity value. The preset reference fluorescence intensity value corresponding to the physical state response unit is retrieved from the initial state database. The reference fluorescence intensity value is the initial fluorescence intensity value recorded by the physical state response unit in the initial state. The difference between the fluorescence intensity value and the preset reference fluorescence intensity value is calculated and determined as the fluorescence intensity attenuation value. The attenuation value characterizes the degree of fluorescence quenching caused by the irreversible chemical color reaction of the fluorescent material in the sensitive medium layer. The fluorescence peak wavelength and fluorescence intensity attenuation value are used as fluorescence characteristic parameters to quantitatively characterize whether an irreversible chemical colorimetric reaction occurs in the sensitive medium layer and the extent of the reaction. Based on the separated interference fringe components, the surface morphology interference fringe variation is extracted, and the interference fringe components are subjected to Fourier transform processing to transform the interference fringe components from the spatial domain to the frequency domain, thereby obtaining the phase distribution information of the interference fringes. The phase distribution information reflects the surface morphology of the micro physical structure array. The spatial distance between adjacent interference fringes is identified from the phase distribution information, that is, the spatial length corresponding to one complete cycle (2π radians) of phase change. This spatial distance is determined as the interference fringe spacing value. The intensity difference between bright and dark fringes in the interference fringes is calculated, that is, the ratio of the average signal intensity of the bright fringe region to the average signal intensity of the dark fringe region is calculated. This ratio is determined as the interference fringe contrast. Retrieve the initial interference fringe spacing value and initial interference fringe contrast value corresponding to the physical state response unit from the initial state database. The initial interference fringe spacing value and initial interference fringe contrast value are the initial interference fringe parameters recorded by the physical state response unit in the initial state. The difference between the currently acquired interference fringe spacing value and the initial interference fringe spacing value is calculated, and this difference is determined as the change in interference fringe spacing. The difference between the currently acquired interference fringe contrast and the initial interference fringe contrast is calculated, and this difference is determined as the change in interference fringe contrast. The change in interference fringe spacing and the change in interference fringe contrast are used as the change in surface morphology interference fringes to quantitatively characterize whether the micro-physical structure array deforms due to the expansion of the sensitive medium layer and the degree of deformation. The parsed data is organized according to a preset data structure to form the current optical feature parameter set of the physical state response unit. The current optical feature parameter set includes at least one of the following: reflection spectrum intensity distribution sequence, fluorescence peak wavelength value, fluorescence intensity attenuation value, and surface morphology interference fringe spacing value. The current optical feature parameter set is associated with and stored with the unique spatial code of the physical state response unit. State offset calculation module: Based on the current physical state information, determine the state offset of each physical state response unit. The state offset is used to characterize the degree of irreversible change of its physical state compared to the initial state. In this embodiment of the invention, the state offset calculation module needs to be specifically described. The state offset calculation module extracts the initial optical feature parameter set corresponding to the physical state response unit from the initial state database based on the unique spatial coding. The initial optical feature parameter set includes the initial reflectance spectrum intensity distribution sequence, the initial fluorescence peak wavelength value, the initial fluorescence intensity value, the initial interference fringe spacing value, and the initial interference fringe contrast value. Based on the current optical feature parameter set and the initial optical feature parameter set, a parameter-by-parameter comparison is performed to generate feature difference values ​​in at least one dimension. The parameter-by-parameter comparison includes: comparing the current reflectance spectral intensity distribution sequence with the initial reflectance spectral intensity distribution sequence to form a reflectance spectral difference sequence; comparing the current fluorescence peak wavelength value with the initial fluorescence peak wavelength value to generate a fluorescence peak wavelength shift; comparing the current fluorescence intensity value with the initial fluorescence intensity value to generate a fluorescence intensity attenuation value; comparing the current interference fringe spacing value with the initial interference fringe spacing value to generate an interference fringe spacing change; and comparing the current interference fringe contrast with the initial interference fringe contrast to generate an interference fringe contrast change. Based on the feature difference value of at least one dimension, the comprehensive state offset of the physical state response unit is determined. The process of determining the comprehensive state offset includes normalizing the feature difference value of each dimension to obtain the normalized change degree, multiplying the normalized change degree of each dimension by a preset weight coefficient to obtain the weighted change degree, and summing the weighted change degrees of all dimensions to obtain the comprehensive state offset.

[0019] After completing the acquisition and storage of the current optical feature parameter set, the unique spatial code of the physical state response unit currently being processed is obtained. This unique spatial code is a unique identifier that distinguishes different physical state response units. It is obtained by decoding the micro physical structure array on the surface of the physical state response unit. The unique spatial code is used as the retrieval index to perform matching queries in the initial state database. The initial state database uses the unique spatial code as the primary key and pre-stores the initial optical feature parameter set of each physical state response unit at the time of deployment. The initial optical feature parameter set includes the initial reflectance spectral intensity distribution sequence, the initial fluorescence peak wavelength value, the initial fluorescence intensity value, the initial interference fringe spacing value, and the initial interference fringe contrast value. Perform precise matching based on the unique spatial code, locate the data record corresponding to the code, and extract all the initial optical feature parameters from the data record as the basis for subsequent comparisons. Based on the current optical feature parameter set and the initial optical feature parameter set, a parameter-by-parameter comparison is performed to generate feature difference values ​​in at least one dimension. Each parameter in the current optical feature parameter set is compared with the corresponding parameter in the initial optical feature parameter set one by one. The comparison process is performed sequentially according to a preset parameter comparison order. For the reflectance spectrum dimension, the current reflectance spectrum intensity distribution sequence is compared with the initial reflectance spectrum intensity distribution sequence wavelength by wavelength. The difference between the current reflectance value and the initial reflectance value is calculated at each wavelength position. The differences at all wavelength positions are arranged in wavelength order to form a reflectance spectrum difference sequence. For the fluorescence spectrum dimension, the current fluorescence peak wavelength value is compared with the initial fluorescence peak wavelength value, and the wavelength shift between the two is calculated. At the same time, the current fluorescence intensity value is compared with the initial fluorescence intensity value, and the difference between the two is calculated as the fluorescence intensity decay value. For the interference fringe dimension, the current interference fringe spacing value is compared with the initial interference fringe spacing value, and the difference between the two is calculated as the spacing change. The current interference fringe contrast is compared with the initial interference fringe contrast, and the difference between the two is calculated as the contrast change. Based on the feature difference value of at least one dimension, the comprehensive state offset of the physical state response unit is determined. The generated feature difference values ​​of multiple dimensions are comprehensively processed to obtain a single quantitative value that can comprehensively characterize the degree of irreversible change of the physical state of the physical state response unit. The comprehensive processing is executed according to the preset fusion rule. The preset fusion rule is based on the sensitivity and reliability of the feature difference values ​​of each dimension to the risk response. First, the feature difference values ​​of each dimension are normalized, that is, the feature difference values ​​of each dimension are mapped to a unified numerical range to eliminate the numerical differences between different dimensions due to different physical meanings and different units. The specific method of normalization is as follows: For each dimension, the maximum possible range of change of that dimension is preset, and the ratio of the actual feature difference value of that dimension to the maximum possible range of change is used as the normalized change degree of that dimension. A preset weight coefficient is assigned to the normalized change degree of each dimension. The weight coefficient is preset according to the response sensitivity and specificity of each dimension to different types of risk mutation factors. The normalized degree of change of each dimension is multiplied by its corresponding weight coefficient to obtain the weighted degree of change of each dimension. Finally, the weighted degrees of change of all dimensions are summed and the sum is determined as the comprehensive state offset of the physical state response unit. The comprehensive state offset is a dimensionless value. The larger the value, the greater the degree of change of the physical state of the physical state response unit compared with the initial state. That is, the stronger the cumulative effect of the risk-induced mutation factors it is exposed to, and the more significant the degree of irreversible change.

[0020] Risk analysis module: Based on the state offset, determine the risk assessment result, which includes the risk type, risk level, and risk location.

[0021] In this embodiment of the invention, the risk analysis module needs to be specifically described. The risk analysis module obtains the comprehensive state offset of multiple physical state response units and their corresponding unique spatial codes, and maps the comprehensive state offset of each physical state response unit to the three-dimensional coordinate position indicated by the unique spatial code to form a risk offset spatial distribution map. Based on the risk offset spatial distribution map, spatial clustering analysis is performed. The comprehensive state offset is compared with the first preset threshold. Spatial coordinate points with a comprehensive state offset greater than or equal to the first preset threshold are marked as risk points, and spatial coordinate points with a comprehensive state offset less than the first preset threshold are marked as non-risk points. Through spatial connectivity judgment, interconnected risk points are merged into continuous regions, and the continuous regions formed by physical state response units with comprehensive state offsets exceeding the first preset threshold are determined. Based on a continuous region, the length, width, and height coordinates of all risk points within the region are obtained. The length, width, and height coordinates are summed to obtain the total length coordinates, width coordinates, and height coordinates, respectively. The total number of risk points is counted. The average length coordinate is obtained by dividing the total length coordinate by the total number of risk points. The average width coordinate is obtained by dividing the total width coordinate by the total number of risk points. The average height coordinate is obtained by dividing the total height coordinate by the total number of risk points. The three-dimensional coordinate point formed by the average length coordinate, average width coordinate, and average height coordinate is determined as the geometric center of the continuous region. The geometric center is used as the risk location. Based on a continuous region, the comprehensive state offset values ​​of risk points within the continuous region are obtained. The minimum, maximum, and average values ​​of the comprehensive state offset are determined. The average or maximum value is compared with a preset risk level classification rule to determine the numerical range to which the average or maximum value belongs. The risk level corresponding to the numerical range is determined as the risk level of the continuous region. The feature difference values ​​of at least one dimension of at least one physical state response unit within the continuous region are obtained. The feature difference values ​​of at least one dimension include at least one of the following: reflectance spectrum difference sequence, fluorescence peak wavelength offset, fluorescence intensity attenuation value, interference fringe spacing change, and interference fringe contrast change. Based on feature difference values ​​in at least one dimension, feature difference value combination types are formed by combining them in a preset order. The feature difference value combination types are compared with the pre-defined correspondence between feature difference value combination types and risk types. The risk types that match the feature difference value combination types are determined as the risk types of continuous regions.

[0022] It should be noted that the comprehensive state offset of all physical state response units is obtained, and each comprehensive state offset is associated with its corresponding unique spatial code and stored. The unique spatial code contains the length coordinate value, width coordinate value and height coordinate value of the physical state response unit in the three-dimensional coordinate system of the grain pile. Using the three-dimensional coordinate position indicated by the unique spatial code as the spatial positioning reference, the comprehensive state offset of each physical state response unit is mapped to the spatial coordinate point. The specific mapping process is as follows: For each physical state response unit, the system reads its unique spatial code and decodes it to obtain the length coordinate value, width coordinate value and height coordinate value of the unit. Using three-dimensional coordinates as spatial locations, the comprehensive state offset of the physical state response unit is used as the attribute value at that location. According to the spatial coordinate arrangement order of all physical state response units, each coordinate location and its corresponding comprehensive state offset are organized into a three-dimensional spatial data matrix. The three-dimensional spatial data matrix is ​​the risk offset spatial distribution map. Based on the risk offset spatial distribution map, spatial clustering analysis is performed to determine the continuous region formed by physical state response units whose comprehensive state offset exceeds the first preset threshold, and to obtain the preset first preset threshold. The first preset threshold is a critical value of comprehensive state offset set in advance according to the requirements of safe storage of grain piles, which is used to distinguish between normal fluctuations and abnormal risks. Traverse each spatial coordinate point in the risk offset spatial distribution map, read the comprehensive state offset value at the coordinate point, compare the value with the first preset threshold, and mark the spatial coordinate point with the comprehensive state offset greater than or equal to the first preset threshold as a risk point. For spatial coordinate points whose overall state offset is less than the first preset threshold, they are marked as non-risk points. After all marking is completed, spatial connectivity judgment is performed to identify which of the coordinate points marked as risk points are spatially adjacent to each other. The judgment rule for spatial adjacency is: if the distance between two risk points in the length coordinate direction, width coordinate direction, or height coordinate direction does not exceed one grid spacing, then the two are determined to be adjacent. All interconnected risk points are grouped into the same risk area to form at least one risk area. From all risk areas, a continuous area composed of risk points is identified. The continuous area is the area in which physical state response units with a comprehensive state offset exceeding a first preset threshold are continuously distributed in space, representing the actual impact range of a risk event occurring inside the grain pile. Based on the continuous region, the geometric center of the continuous region is determined as the risk location. The spatial coordinates of all risk points contained in the continuous region are obtained. For each risk point, its length coordinate value, width coordinate value and height coordinate value are extracted. The length coordinate value, width coordinate value and height coordinate value are summed respectively to obtain the sum of length coordinates, the sum of width coordinates and the sum of height coordinates. The total number of risk points within a continuous area is counted. The sum of the length coordinates is divided by the total number of risk points to obtain the average length coordinate. The sum of the width coordinates is divided by the total number of risk points to obtain the average width coordinate. The sum of the height coordinates is divided by the total number of risk points to obtain the average height coordinate. The three-dimensional coordinate point formed by the average length coordinate, the average width coordinate, and the average height coordinate is determined as the geometric center of the continuous area. The geometric center represents the concentrated location of risk events in the grain pile space. Based on the numerical distribution range of the comprehensive state offset of the physical state response unit in a continuous area, the risk level is determined, the comprehensive state offset values ​​of all risk points in the continuous area are obtained, the minimum, maximum and average values ​​of the comprehensive state offset in the continuous area are determined, and the preset risk level classification rules are obtained. The risk level classification rules divide the numerical range of the comprehensive state offset into multiple continuous intervals, each interval corresponding to a risk level. For example, the first numerical interval corresponds to level one risk, the second numerical interval corresponds to level two risk, and the third numerical interval corresponds to level three risk. The larger the value, the higher the risk level. The average value of the comprehensive state offset within a continuous region is compared with the preset risk level classification rules to determine the numerical range to which the average value belongs. The risk level corresponding to the numerical range is determined as the risk level of the continuous region. The maximum value of the comprehensive state offset within the continuous region is compared with the preset risk level classification rules to determine the risk level corresponding to the numerical range to which the maximum value belongs. The feature difference value of at least one dimension of at least one physical state response unit in a continuous region is obtained. From the determined continuous region, at least one physical state response unit is selected as a sampling unit. The sampling unit can be the physical state response unit with the largest comprehensive state offset in the continuous region, or the physical state response unit at the geometric center of the continuous region, or the set of all physical state response units in the continuous region. For each selected physical state response unit, the feature difference value of at least one dimension of the physical state response unit is extracted from the intermediate data generated during its state offset calculation. The feature difference values ​​in at least one dimension include at least one of the following: reflectance spectrum difference sequence, fluorescence peak wavelength shift, fluorescence intensity attenuation value, interference fringe spacing change, and interference fringe contrast change. These feature difference values ​​record the specific changes of the physical state response unit in different physical attribute dimensions. Different types of risk mutagenic factors will cause changes in the combination of feature difference values ​​in different dimensions.

[0023] Based on the combination type of feature difference values ​​in at least one dimension, the risk type corresponding to the combination type is determined, and a preset correspondence between the feature difference value combination type and the risk type is obtained. This correspondence is obtained through pre-test calibration. During the pre-calibration process, simulation experiments are conducted on known risk types such as grain mold, pest infestation, localized heating, and abnormal fermentation. The combination type of feature difference values ​​generated by each physical state response unit when each risk occurs is recorded, establishing a unique correspondence between the feature difference value combination type and the risk type. For example, a combination type with significant changes in the reflectance spectrum difference sequence in a specific band, large fluorescence intensity attenuation, and small changes in interference fringes corresponds to grain mold risk; a combination type with significant fluorescence peak wavelength shift, large changes in interference fringe spacing, and large changes in contrast corresponds to pest infestation risk. Combinations of reflectance spectral difference sequences that exhibit uniform variation across the entire wavelength band, significant variation in interference fringe spacing, and relatively small variation in fluorescence parameters correspond to localized heating risks. The acquired feature difference values ​​of at least one dimension are combined in a preset order to form feature difference value combination types. These combination types are then compared with pre-defined correspondences to find risk types that match them. If a match is found, the risk type is identified as the risk type for that continuous region. If a match fails, the system outputs an unrecognizable risk type identifier.

[0024] All preset thresholds, preset numerical ranges, preset weight coefficients, and preset correspondences involved in this application are obtained through pre-calibration. The pre-calibration specifically includes: collecting the comprehensive state offset distribution data of each physical state response unit under normal storage conditions without risk event interference; determining the upper limit of the normal fluctuation range as the first preset threshold through statistical analysis of the distribution data; conducting simulation experiments for different types of known risk events; collecting the comprehensive state offset data and feature difference value data of each dimension of each physical state response unit during the simulation experiments; dividing the numerical range corresponding to each risk level by statistically analyzing the numerical distribution of the comprehensive state offset under different risk levels; establishing the correspondence between the combination type of feature difference value and the risk type by statistically analyzing the combination type of feature difference value under different risk types; and determining the preset weight coefficient corresponding to each dimension by analyzing the contribution of each dimension's feature difference value in the calculation of the comprehensive state offset.

[0025] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0026] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0027] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0028] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0029] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of protection of the described technical solution.

Claims

1. An intelligent safety risk monitoring and early warning management system for the grain industry, characterized in that, include: Sensor deployment module: Several physical state response units are deployed within the grain pile, and the physical state response units are composed of physically variable materials that are sensitive to risk-induced mutation factors; Data acquisition module: Based on a non-contact scanning device, it performs a full-domain scan of several physical state response units to obtain the current physical state information of each physical state response unit; State offset calculation module: Based on the current physical state information, determine the state offset of each physical state response unit. The state offset is used to characterize the degree of irreversible change of its physical state compared to the initial state. Risk analysis module: Based on the state offset, determine the risk assessment result, which includes the risk type, risk level, and risk location.

2. The intelligent safety risk monitoring and early warning management system for the grain industry according to claim 1, characterized in that, The sensor deployment module deploys physical state response units in the form of three-dimensional grid nodes inside the grain pile; Based on the deployment strategy, a unique spatial code is generated for each physical state response unit, and the unique spatial code is associated with its three-dimensional coordinate information in the grain pile; Based on unique spatial coding, an initial state database is established. The initial state database is used to record the initial physical state parameters of each physical state response unit during deployment.

3. The intelligent safety risk monitoring and early warning management system for the grain industry according to claim 1, characterized in that, The data acquisition module acquires a broadband excitation beam emitted by a non-contact scanning device. The broadband excitation beam sequentially covers the spatial location of each physical state response unit based on a preset scanning path, and the spot diameter of the broadband excitation beam matches the size of the physical state response unit. Based on a broadband excitation beam, optical excitation is performed on the surface of each physical state response unit to obtain the feedback signal generated by each physical state response unit. The feedback signals are received sequentially by the optical receiver built into the non-contact scanning device according to the preset scanning path, and each feedback signal is associated with and stored with the current scanning position information.

4. The intelligent safety risk monitoring and early warning management system for the grain industry according to claim 3, characterized in that, Based on the feedback signal, the current optical characteristic parameter set of each physical state response unit is determined. The current optical characteristic parameter set is obtained by analyzing the feedback signal. The analysis process includes extracting at least one of the following from the feedback signal: the reflection spectrum intensity distribution sequence, the fluorescence peak wavelength value, the fluorescence intensity attenuation value, and the surface morphology interference fringe spacing value. The analyzed data is then associated and stored with the unique spatial code of the physical state response unit.

5. The intelligent safety risk monitoring and early warning management system for the grain industry according to claim 1, characterized in that, The state offset calculation module extracts the initial optical feature parameter set corresponding to the physical state response unit from the initial state database based on a unique spatial code. The initial optical feature parameter set includes the initial reflectance spectrum intensity distribution sequence, the initial fluorescence peak wavelength value, the initial fluorescence intensity value, the initial interference fringe spacing value, and the initial interference fringe contrast value. Based on the current optical feature parameter set and the initial optical feature parameter set, a parameter-by-parameter comparison is performed to generate feature difference values ​​in at least one dimension. The parameter-by-parameter comparison includes: comparing the current reflectance spectral intensity distribution sequence with the initial reflectance spectral intensity distribution sequence to form a reflectance spectral difference sequence; comparing the current fluorescence peak wavelength value with the initial fluorescence peak wavelength value to generate a fluorescence peak wavelength offset; comparing the current fluorescence intensity value with the initial fluorescence intensity value to generate a fluorescence intensity attenuation value; comparing the current interference fringe spacing value with the initial interference fringe spacing value to generate an interference fringe spacing change; and comparing the current interference fringe contrast with the initial interference fringe contrast to generate an interference fringe contrast change.

6. The intelligent safety risk monitoring and early warning management system for the grain industry according to claim 5, characterized in that, Based on the feature difference value of at least one dimension, the comprehensive state offset of the physical state response unit is determined. The process of determining the comprehensive state offset includes normalizing the feature difference value of each dimension to obtain the normalized change degree, multiplying the normalized change degree of each dimension by a preset weight coefficient to obtain the weighted change degree, and summing the weighted change degrees of all dimensions to obtain the comprehensive state offset.

7. The intelligent safety risk monitoring and early warning management system for the grain industry according to claim 1, characterized in that, The risk analysis module obtains the comprehensive state offset of multiple physical state response units and their corresponding unique spatial codes, and maps the comprehensive state offset of each physical state response unit to the three-dimensional coordinate position indicated by the unique spatial code to form a risk offset spatial distribution map. Based on the risk offset spatial distribution map, spatial clustering analysis is performed. The comprehensive state offset is compared with a first preset threshold. Spatial coordinate points with a comprehensive state offset greater than or equal to the first preset threshold are marked as risk points, and spatial coordinate points with a comprehensive state offset less than the first preset threshold are marked as non-risk points. Through spatial connectivity judgment, interconnected risk points are merged into continuous regions, and the continuous regions formed by physical state response units with comprehensive state offsets exceeding the first preset threshold are determined.

8. The intelligent safety risk monitoring and early warning management system for the grain industry according to claim 7, characterized in that, Based on a continuous region, obtain the length, width, and height coordinates of all risk points within the region. Sum the length, width, and height coordinates to obtain the total length, width, and height coordinates, respectively. Count the total number of risk points. Divide the total length coordinates by the total number of risk points to obtain the average length coordinate. Divide the total width coordinates by the total number of risk points to obtain the average width coordinate. Divide the total height coordinates by the total number of risk points to obtain the average height coordinate. Determine the three-dimensional coordinate point formed by the average length, width, and height coordinates as the geometric center of the continuous region, and use the geometric center as the risk location.

9. A smart safety risk monitoring and early warning management system for the grain industry according to claim 7, characterized in that, Based on a continuous region, the comprehensive state offset values ​​of risk points within the continuous region are obtained, the minimum, maximum, and average values ​​of the comprehensive state offset are determined, the average or maximum value is compared with the preset risk level classification rules, the value range to which the average or maximum value belongs is determined, and the risk level corresponding to the value range is determined as the risk level of the continuous region. The feature difference value of at least one dimension of at least one physical state response unit in a continuous region is obtained, wherein the feature difference value of at least one dimension includes at least one of the following: reflectance spectrum difference sequence, fluorescence peak wavelength shift, fluorescence intensity attenuation value, interference fringe spacing change and interference fringe contrast change. Based on feature difference values ​​in at least one dimension, feature difference value combination types are formed by combining them in a preset order. The feature difference value combination types are compared with the pre-defined correspondence between feature difference value combination types and risk types. The risk types that match the feature difference value combination types are determined as the risk types of continuous regions.

Images