CN-121981655-A - Fruit cold chain inventory management method based on Internet of things

CN121981655ACN 121981655 ACN121981655 ACN 121981655ACN-121981655-A

Abstract

The invention relates to the technical field of intelligent cold chain logistics and Internet of things data, in particular to a fruit cold chain inventory management method based on the Internet of things, which comprises the steps of acquiring environment time sequence data, real-time state data of inventory assets and static attribute data in a cold chain inventory environment; the method comprises the steps of constructing a biological thermodynamic lossless model, generating an ideal inventory attenuation curve representing natural loss trend, generating a theoretical fault feature library containing various theoretical fault forms, calculating real residual vectors between real-time state data and the ideal inventory attenuation curve, calculating the similarity of each vector in a real residual vector set and a theoretical residual vector set, judging the loss type of the current inventory, executing a corresponding hierarchical management and control strategy, and completing inventory anomaly tracing and closed loop treatment.

Inventors

Lin Xueruo
ZHANG HUINAN
PAN XIA

Assignees

延安自然搭档农业发展有限公司

Dates

Publication Date: 20260505
Application Date: 20260409

Claims (8)

1. Fruit cold chain inventory management method based on the Internet of things is characterized by comprising the following steps: acquiring environment time sequence data, real-time state data of inventory assets and static attribute data in a cold chain warehousing environment; Based on the environmental time sequence data and the static attribute data, constructing a biological thermodynamic lossless model, calculating theoretical total mass loss of a target stock by using the biological thermodynamic lossless model, and generating an ideal stock attenuation curve representing a natural loss trend; Injecting preset abnormal mode parameters into the biological thermodynamic lossless model to perform multidimensional fault simulation, and generating a theoretical fault feature library containing various theoretical fault forms; Respectively calculating a real residual vector between the real-time state data and an ideal inventory attenuation curve and a theoretical residual vector set between each theoretical fault form in the theoretical fault feature library and the ideal inventory attenuation curve; Calculating the similarity between the actual residual vector and each vector in the theoretical residual vector set, and judging the loss type of the current stock based on the maximum similarity matching result; and responding to the loss type, executing a corresponding hierarchical management and control strategy, and completing inventory anomaly tracing and closed loop treatment.
2. The internet of things-based fruit cold chain inventory management method of claim 1, wherein obtaining environmental time series data, real-time status data of inventory assets, and static attribute data in a cold chain inventory environment comprises: Acquiring environmental time sequence data by an environmental sensor array deployed in a refrigeration house, wherein the environmental time sequence data comprises a temperature and humidity curve, ethylene concentration, oxygen concentration and carbon dioxide concentration; Collecting real-time state data of inventory assets through intelligent bearing equipment and visual perception equipment, wherein the real-time state data comprises real-time weight flow data and surface hyperspectral characteristic data; static attribute data is extracted from a preset database, wherein the static attribute data comprises the respiration consumption rate of stock varieties, standard transpiration coefficients, the initial time of warehouse entry and the initial quality of warehouse entry.
3. The internet of things-based fruit cold chain inventory management method of claim 2, wherein constructing a thermodynamic lossless model based on the environmental time series data and static attribute data, calculating a theoretical total mass loss of a target inventory using the thermodynamic lossless model, generating an ideal inventory decay curve that characterizes natural loss trends, comprises: calculating the water pressure difference of the current environment based on the temperature and humidity curve; Combining the respiratory consumption rate and the standard transpiration coefficient, and calculating the theoretical total mass loss rate in unit time by using a calculation algorithm based on a Pengman-Meng Tesi equation; And carrying out time integration on the theoretical total mass loss rate, and deducting an integration result from the warehouse-in initial mass to generate an ideal inventory attenuation curve which monotonically decreases with time.
4. The method for fruit cold chain inventory management based on the internet of things according to claim 3, wherein the step of injecting preset abnormal mode parameters into the biological thermodynamic lossless model for multidimensional fault simulation based on a knowledge driving strategy, and the step of generating a theoretical fault feature library comprising a plurality of theoretical fault forms comprises the steps of: Extracting a preset theft library transfer factor, constructing a negative step function, and overlapping the negative step function to the ideal inventory attenuation curve to generate a first type of theoretical fault form; extracting a preset equipment fault decay factor, constructing an attenuation coefficient which has an exponential relation with the environmental accumulated temperature, and overlapping the attenuation coefficient with the ideal inventory attenuation curve to generate a second theoretical fault form; Extracting a preset package breakage factor, constructing a linear deviation function with a constant slope, and overlapping the linear deviation function with the constant slope to the ideal inventory attenuation curve to generate a third type of theoretical fault form; And collecting the first type of theoretical fault morphology, the second type of theoretical fault morphology and the third type of theoretical fault morphology, and constructing the theoretical fault feature library.
5. The internet of things-based fruit cold chain inventory management method of claim 4, wherein calculating real residual vectors between the real-time status data and an ideal inventory decay curve, and a set of theoretical residual vectors between each theoretical fault modality in the theoretical fault signature library and an ideal inventory decay curve, respectively, comprises: Performing time domain alignment and differential operation on the real-time weight flow data and the ideal inventory attenuation curve, and calculating the difference value of the real-time weight flow data subtracted by the ideal inventory attenuation curve to be used as a real residual error vector representing unnatural factors; and respectively carrying out differential operation on each theoretical fault form in the theoretical fault feature library and the ideal inventory attenuation curve, and calculating the difference value of subtracting the ideal inventory attenuation curve from each theoretical fault form to serve as a theoretical residual error vector set.
6. The method of claim 5, wherein calculating the similarity between the real residual vector and each vector in the set of theoretical residual vectors, and determining the type of loss of the current inventory based on the maximum similarity matching result comprises: Calculating the similarity value of the actual residual vector and each theoretical residual vector in the theoretical residual vector set by adopting a dynamic time warping algorithm or a cosine similarity algorithm; If the module length of the actual residual vector is smaller than a preset noise threshold value, judging that the loss type of the current stock is normal natural loss; if the modulus of the actual residual vector is greater than or equal to the noise threshold and the maximum value in the similarity value is greater than a preset matching threshold, judging the abnormal mode represented by the theoretical fault form corresponding to the maximum value as the loss type of the current stock; And if the modulus of the actual residual vector is greater than or equal to the noise threshold and the maximum value in the similarity value is less than or equal to the matching threshold, judging that the loss type of the current stock is unknown abnormality.
7. The internet of things-based fruit cold chain inventory management method of claim 6, wherein executing a corresponding hierarchical management and control policy in response to the loss type comprises: If the loss type is judged to be normal natural loss, automatically ignoring weight reduction data, not triggering an alarm and recording the weight reduction data as a compliance physical weight reduction in a system; If the loss type is determined to be abnormal in the theft and the database shift, triggering a security alarm instruction and marking a time stamp of the occurrence of the abnormality; and if the loss type is judged to be the equipment failure and rot abnormality, generating a cold chain equipment maintenance work order, and adjusting the expected shelf life of the inventory assets.
8. The internet of things-based fruit cold chain inventory management method of claim 7, further comprising: storing the judged loss type and the corresponding real residual vector into a historical database; and carrying out iterative correction on the biological thermodynamic lossless model and the abnormal mode parameters by using the stored historical data so as to improve the accuracy of subsequent judgment.

Description

Fruit cold chain inventory management method based on Internet of things Technical Field The invention relates to the technical field of intelligent cold chain logistics and internet of things data, in particular to a fruit cold chain inventory management method based on the internet of things. Background In the current fruit cold chain inventory management scene, inventory assets continuously generate natural weight loss caused by respiration and transpiration along with the storage time, and meanwhile, the risks of abnormal weight reduction caused by unnatural factors such as theft, warehouse moving and decay caused by equipment faults are also faced, and dynamic change characteristics such as defrosting, temperature fluctuation and the like exist in a refrigeration house environment; In order to manage and control inventory loss, the prior proposal generally adopts a monitoring method based on a fixed threshold or a simple linear depreciation model to estimate inventory state, while the proposal has certain inventory capability when the environment is stable, and because the proposal can not dynamically adjust loss standard according to temperature and humidity changes, and can not effectively strip natural biological loss background noise influenced by the environment from real abnormal weight reduction signals, when facing to temperature fluctuation or trace abnormality of a refrigerator, high false alarm rate or false alarm is generated due to incapability of accurately distinguishing compliance physical weight reduction and management error, and fine tracing and closed loop treatment of inventory risk are difficult to realize, therefore, how to combine environment perception data to construct dynamic loss standard, realize accurate decoupling and real-time judgment of natural biological loss and abnormal management loss becomes a technical problem to be solved. Disclosure of Invention In order to solve the technical problems, the invention provides a fruit cold chain inventory management method based on the Internet of things, and specifically, the technical scheme of the invention comprises the following steps: acquiring environment time sequence data, real-time state data of inventory assets and static attribute data in a cold chain warehousing environment; Based on the environmental time sequence data and the static attribute data, constructing a biological thermodynamic lossless model, calculating theoretical total mass loss of a target stock by using the biological thermodynamic lossless model, and generating an ideal stock attenuation curve representing a natural loss trend; Injecting preset abnormal mode parameters into the biological thermodynamic lossless model to perform multidimensional fault simulation, and generating a theoretical fault feature library containing various theoretical fault forms; Respectively calculating a real residual vector between the real-time state data and an ideal inventory attenuation curve and a theoretical residual vector set between each theoretical fault form in the theoretical fault feature library and the ideal inventory attenuation curve; Calculating the similarity between the actual residual vector and each vector in the theoretical residual vector set, and judging the loss type of the current stock based on the maximum similarity matching result; and responding to the loss type, executing a corresponding hierarchical management and control strategy, and completing inventory anomaly tracing and closed loop treatment. Optionally, acquiring the environmental time sequence data, the real-time status data of the inventory asset and the static attribute data in the cold chain warehousing environment includes: Acquiring environmental time sequence data by an environmental sensor array deployed in a refrigeration house, wherein the environmental time sequence data comprises a temperature and humidity curve, ethylene concentration, oxygen concentration and carbon dioxide concentration; Collecting real-time state data of inventory assets through intelligent bearing equipment and visual perception equipment, wherein the real-time state data comprises real-time weight flow data and surface hyperspectral characteristic data; static attribute data is extracted from a preset database, wherein the static attribute data comprises the respiration consumption rate of stock varieties, standard transpiration coefficients, the initial time of warehouse entry and the initial quality of warehouse entry. Optionally, constructing a thermodynamic lossless model based on the environmental time sequence data and the static attribute data, calculating a theoretical total mass loss of the target inventory using the thermodynamic lossless model, and generating an ideal inventory decay curve representing a natural loss trend includes: calculating the water pressure difference of the current environment based on the temperature and humidity curve; Combining the respiratory consumption ra