CN116663753B - Cold chain food distribution prediction method and system - Google Patents
Cold chain food distribution prediction method and system Download PDFInfo
- Publication number
- CN116663753B CN116663753B CN202310953977.4A CN202310953977A CN116663753B CN 116663753 B CN116663753 B CN 116663753B CN 202310953977 A CN202310953977 A CN 202310953977A CN 116663753 B CN116663753 B CN 116663753B
- Authority
- CN
- China
- Prior art keywords
- vector
- parameter
- information
- node
- parameter vector
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 57
- 235000013305 food Nutrition 0.000 title claims abstract description 43
- 239000013598 vector Substances 0.000 claims abstract description 400
- 238000004458 analytical method Methods 0.000 claims abstract description 128
- 238000012545 processing Methods 0.000 claims abstract description 27
- 238000004364 calculation method Methods 0.000 claims abstract description 23
- 230000000630 rising effect Effects 0.000 claims abstract description 10
- 238000005057 refrigeration Methods 0.000 claims description 45
- 238000012384 transportation and delivery Methods 0.000 claims description 26
- 238000006243 chemical reaction Methods 0.000 claims description 17
- 238000011161 development Methods 0.000 claims description 10
- 230000008859 change Effects 0.000 claims description 8
- 238000013139 quantization Methods 0.000 claims description 8
- 230000001174 ascending effect Effects 0.000 claims description 7
- 238000005192 partition Methods 0.000 claims description 6
- 238000013480 data collection Methods 0.000 claims description 4
- 238000000605 extraction Methods 0.000 claims description 4
- 230000008569 process Effects 0.000 description 19
- 230000006870 function Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 3
- 238000005457 optimization Methods 0.000 description 3
- 230000007613 environmental effect Effects 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 238000013178 mathematical model Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 235000013611 frozen food Nutrition 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000013439 planning Methods 0.000 description 1
- 238000004321 preservation Methods 0.000 description 1
- 238000011002 quantification Methods 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 238000010257 thawing Methods 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
- 238000012800 visualization Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
- G06Q10/047—Optimisation of routes or paths, e.g. travelling salesman problem
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
- G06Q10/083—Shipping
- G06Q10/0835—Relationships between shipper or supplier and carriers
- G06Q10/08355—Routing methods
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Physics & Mathematics (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Development Economics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Game Theory and Decision Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- General Preparation And Processing Of Foods (AREA)
Abstract
The invention relates to a cold chain food distribution prediction method and a system, wherein a merging and sorting unit carries out merging calculation on first parameter information by acquiring the first parameter information input by a sensing unit and carries out vectorization by combining a vector machine to obtain a first parameter vector; the first parameter vector is placed in a two-dimensional coordinate by adopting a node analysis network to carry out node analysis, so that the first parameter vector is identified as gradient rising or gradient falling in the two-dimensional coordinate; acquiring second parameter information input by the sensing unit; generating a matched prediction result based on the first parameter vector and the second parameter vector by using a preset long and short time sequence prediction unit; the optimal control of the distribution time and the distribution path is realized through the accurate processing and analysis of humidity, temperature, volume, route, time and heat parameters, the distribution precision and efficiency are improved, and the food loss and the distribution cost are reduced.
Description
Technical Field
The invention relates to the technical field of food distribution, in particular to a cold chain food distribution prediction method and a cold chain food distribution prediction system.
Background
The cold chain food distribution industry is a ubiquitous and growing industry to meet the demands of consumers for fresh foods, and various companies put more effort in the distribution link to ensure the freshness and safety of the foods, wherein the convenience and punctual rate of cold chain distribution are more required, the quality of the foods is damaged due to overlong distribution time, but factors such as climate, traffic conditions and the like can influence the efficiency and punctual rate of cold chain distribution.
Disclosure of Invention
The invention mainly aims to provide a cold chain food distribution prediction method and system, which realize the optimal control of distribution time and paths, improve distribution precision and efficiency and reduce food loss and distribution cost through the accurate processing and analysis of parameters such as humidity, temperature, volume, route, time, heat and the like.
In order to achieve the above object, the present invention provides a cold chain food distribution prediction method, comprising the steps of:
acquiring first parameter information input by a sensing unit, wherein the first parameter information comprises, but is not limited to, humidity information and temperature information;
the method comprises the steps of carrying out merging calculation on first parameter information based on a preset merging and sorting unit, and carrying out vectorization on the first parameter information after merging calculation by combining a vector machine arranged in the merging and sorting unit to obtain a first parameter vector;
the first parameter vector is placed in a two-dimensional coordinate by adopting a preset node analysis network to carry out node analysis, so that the first parameter vector is identified as gradient rising or gradient falling in the two-dimensional coordinate;
acquiring second parameter information input by the sensing unit, wherein the second parameter information comprises, but is not limited to, volume information, route information, time information and heat information;
And carrying out vectorization on second parameter information by using an input gate of a preset long and short time sequence prediction unit, inputting the second parameter vector generated after vectorization into a node analysis network, calibrating a first parameter vector and a second parameter vector in the node analysis network by using the time information, generating a corresponding journey progress vector by using distance data in the route information, loading the journey progress vector into the node analysis network, identifying vector offset of the journey progress vector on the node and the second parameter vector by using a cosine prediction algorithm, correspondingly adjusting node information of the journey progress vector on the node analysis network by using the vector offset, and generating a matched prediction result based on the node information.
Further, the step of performing merging calculation on the first parameter information based on a preset merging and sorting unit, and performing vectorization on the first parameter information after merging calculation by combining a vector machine arranged in the merging and sorting unit to obtain a first parameter vector includes:
carrying out numerical processing on the first parameter information, and carrying out data collection on the numerical first parameter information through a merging and sorting unit to obtain parameter information with a digital group based on a time sequence collection;
Carrying out vector conversion on first parameter information carrying a digital group by using a numerical value vector conversion method preset in a vector machine so as to generate a first parameter vector;
the method for digitally processing the first parameter information comprises the following steps:
performing information calibration on the humidity information and the temperature information of the first parameter information based on a time stamp;
and combining the values in the calibrated humidity information and the calibrated temperature information based on the time stamp to obtain a parameter information integrated with a digital group.
Further, the step of performing vector conversion on the first parameter information carrying the digital group by using a numerical vector conversion method preset in the vector machine to generate a first parameter vector includes:
identifying corresponding humidity numbers and temperature numbers in a plurality of digital groups based on a quality quantization line, wherein the humidity numbers and the temperature numbers are numerical information of humidity information and temperature information respectively;
and vectorizing the quality quantization lines carrying humidity numbers and temperature numbers respectively and performing reverse processing to obtain a first parameter vector preset as gradient descent after reverse processing.
Further, the step of placing the first parameter vector in a two-dimensional coordinate by using a preset node analysis network to perform node analysis so as to identify the first parameter vector as gradient rising or gradient falling in the two-dimensional coordinate includes:
Identifying a current time stamp in real time, and correspondingly updating an abscissa in a preset two-dimensional coordinate based on the current time stamp;
adding the first parameter vector on a two-dimensional coordinate based on a current time stamp, generating a plurality of nodes through refrigerating property preset on an ordinate, and correspondingly marking the numerical value of the first parameter vector based on the numerical values of the plurality of nodes so as to set the first parameter vector in a node analysis network;
carrying out refrigeration numerical analysis on the marked node numerical values and the first parameter vector numerical values to judge the numerical values of the first parameter vector and a preset suitable refrigeration threshold value so as to judge whether the first parameter vector accords with refrigeration or does not accord with refrigeration;
if the first parameter vector accords with refrigeration, the first parameter vector is in gradient ascending in the node analysis network, and otherwise, the first parameter vector is in gradient descending.
Further, wherein: the refrigerating performance comprises four items of high-quality refrigerating conditions, common refrigerating conditions, inferior refrigerating conditions and non-conforming refrigerating conditions, the four items are correspondingly defined by preset numerical values, each refrigerating performance belongs to a numerical value interval, and the step of generating a plurality of nodes through the refrigerating performance preset by the ordinate comprises the following steps:
Identifying a humidity number and a temperature number carried by a first parameter vector, and judging a corresponding interval of four refrigeration items in which the humidity number and the temperature number are located;
the first parameter vector is added with a plurality of sets of humidity numbers and temperature numbers according to the current time stamp along with the change of the current time stamp, and the humidity numbers and the temperature numbers of the sets are identified in real time and are changed on a coordinate system.
Further, the step of vectorizing the second parameter information by using an input gate of the preset long and short time sequence prediction unit to input the second parameter vector generated after vectorizing to the node analysis network includes:
extracting characteristic information in the second parameter information by using an input gate of the long and short time sequence prediction unit, wherein the characteristic information comprises digital information corresponding to volume, route, time and heat;
carrying out partition vectorization on digital information corresponding to the volume, the route, the time and the heat after feature extraction to generate a plurality of parameter factors, and generating a time vector based on the plurality of parameter factors;
the time vectors are all loaded into the node analysis network as second parameter vectors, and the time vectors carry volume factors, route factors, time factors and heat factors.
Further, the step of calibrating the first parameter vector and the second parameter vector in the node analysis network through the time information includes:
determining a starting node in a node analysis network based on a time factor in the second parameter vector;
generating a period of prediction time based on the time factor and the route factor in the second parameter vector on the premise that the abscissa is time;
determining the refrigeration of the ordinate of the matched node analysis network based on the volume factor and the heat factor;
and determining cold storage property through the starting node, the predicted time period and the volume factor and the heat factor to generate a second parameter vector in the node analysis network.
Further, the step of generating a corresponding travel progress vector according to the distance data in the route information and loading the travel progress vector into a node analysis network includes:
a corresponding distance schedule is called through the distance data, so that travel schedule data corresponding to the travel schedule vector are correspondingly obtained from the distance schedule, the travel schedule vector is obtained after the travel schedule data are vectorized, and the travel schedule vector is used as a preset distribution standard;
And loading the travel progress vector into a node analysis network, wherein the starting node of the travel progress vector is consistent with the starting node of the second parameter vector.
Further, the step of identifying the vector offset of the travel progress vector and the second parameter vector on the node by adopting a cosine prediction algorithm, correspondingly adjusting the node information of the travel progress vector on the node analysis network through the vector offset, and generating a matched prediction result based on the node information comprises the following steps:
identifying an included angle between the travel progress vector and the second parameter vector in the node analysis network, wherein the included angle is only confirmed from the vector line of the travel progress vector and the second parameter vector, and the obtained included angle is a vector offset;
identifying each offset node generating a vector offset, and calculating the offset node and the vector offset by a cosine prediction algorithm to calculate the cosine development direction of the offset node;
based on the cosine development direction and combined with the refrigerating property and the predicted time period of the second parameter vector, correspondingly generating terminal node data of the predicted time period, wherein the terminal node data comprises a delivery time and a refrigerating property value;
And correspondingly adjusting the travel progress vector based on the final node data, and displaying the travel progress vector adjusted in real time as a prediction result.
The invention also provides a cold chain food distribution prediction system, which comprises:
a first acquisition unit for acquiring first parameter information input by the sensing unit, wherein the first parameter information comprises, but is not limited to, humidity information and temperature information;
the merging unit is used for merging and calculating the first parameter information based on a preset merging and sorting unit, and vectorizing the first parameter information after merging and calculating by combining a vector machine arranged in the merging and sorting unit to obtain a first parameter vector;
the node network unit is used for placing the first parameter vector in a two-dimensional coordinate by adopting a preset node analysis network to perform node analysis so as to identify that the first parameter vector is in gradient ascending or gradient descending in the two-dimensional coordinate;
a second acquisition unit for acquiring second parameter information input by the sensing unit, wherein the second parameter information comprises, but is not limited to, volume information, route information, time information and heat information;
the model processing unit is used for vectorizing the second parameter information by utilizing an input gate of the preset long and short time sequence prediction unit, inputting the second parameter vector generated after vectorizing into the node analysis network, calibrating the first parameter vector and the second parameter vector in the node analysis network by the time information, generating a corresponding travel progress vector by distance data in the route information, loading the corresponding travel progress vector into the node analysis network, identifying vector offset of the travel progress vector and the second parameter vector on the node by adopting a cosine prediction algorithm, correspondingly adjusting node information of the travel progress vector on the node analysis network by the vector offset, and generating a matched prediction result based on the node information.
The cold chain food distribution prediction method and system provided by the invention have the following beneficial effects:
(1) Enhancing precision: the parameters are processed by adopting the merging and sorting unit and the vector machine, so that the input humidity information and temperature information can be analyzed and understood more accurately, and the prediction accuracy is improved.
(2) Enhanced understandability and visualization: by carrying out node analysis in the two-dimensional coordinates, whether the parameter vector is ascending or descending can be conveniently identified, so that the state transition and trend of the data are more visualized, and the understanding and tracking are facilitated.
(3) Increasing the multi-element data control capability: when the second parameter information is processed, the multi-element data such as volume information, route information, time information, heat information and the like are integrated, so that the prediction is more in line with the actual scene, and the control and management effects of cold chain food distribution are improved.
(4) And the dynamic adaptability is improved: the node analysis network is used, so that the prediction result can adapt to various dynamic change conditions, the first parameter vector and the second parameter vector are calibrated, and a matched prediction result is generated, so that the prediction effect can be adjusted and optimized no matter how the environmental condition changes.
(5) Optimizing path and time management: according to the method, the vector offset of the travel progress vector and the second parameter vector on the node is identified through a cosine prediction algorithm, the travel progress can be adjusted according to the prediction result, and the food quality is ensured while the distribution time and the distribution path are maximized and optimized.
(6) The distribution efficiency is improved: by applying the prediction method, the key information such as the time, the environment requirement, the travel path and the like of the cold chain food distribution can be accurately predicted, the distribution efficiency can be greatly improved by the optimization effect, and meanwhile, the food loss and the distribution cost are reduced.
Drawings
FIG. 1 is a schematic diagram of a method of cold chain food delivery prediction in accordance with an embodiment of the present invention;
FIG. 2 is a block diagram of a cold chain food delivery prediction system in accordance with one embodiment of the present invention;
the achievement of the objects, functional features and advantages of the present invention will be further described with reference to the accompanying drawings, in conjunction with the embodiments.
Detailed Description
The present invention will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
Referring to fig. 1, a cold chain food delivery prediction method includes the steps of:
s1, acquiring first parameter information input by a sensing unit, wherein the first parameter information comprises, but is not limited to, humidity information and temperature information;
in S1, the system needs to collect first parameter information input by the sensor. This information is primarily related to the environmental factors closely related to cold chain food distribution-humidity and temperature. Temperature and humidity are two very critical parameters during the cold chain delivery process. The temperature and humidity requirements for different foods are different, which requires real-time acquisition by sensors. For example, for frozen foods, it is desirable to ensure that thawing does not occur for a short period of time during the dispensing process; for raw fresh food, too low or too high a humidity may affect the quality of the food.
S2, carrying out merging calculation on the first parameter information based on a preset merging and sorting unit, and carrying out vectorization on the first parameter information after merging calculation by combining a vector machine arranged in the merging and sorting unit to obtain a first parameter vector;
in S2, a preset merging and sorting unit is adopted to carry out merging calculation on the humidity and temperature information. Merge sort is a very efficient sort algorithm that first breaks the dataset into smaller portions, sorts each of the smaller portions, and then merges the portions together to form a complete and sorted set of data. In this process, the humidity and temperature information will be uniformly ordered and processed. And secondly, vectorizing the processed humidity and temperature information based on a set vector machine. Vectorization is the mapping of humidity and temperature information into a vector space of high dimensions, so that this information can be used in mathematical models or machine learning algorithms for subsequent calculations and predictions. Through the above steps we have obtained a vector representation of humidity and temperature information-the first parameter vector. This vector will be used in subsequent predictions and analyses.
S3, placing the first parameter vector in a two-dimensional coordinate by adopting a preset node analysis network to perform node analysis so as to identify that the first parameter vector is gradient rising or gradient falling in the two-dimensional coordinate;
in S3, the node analysis network is a deep learning network that can analyze highly complex and interrelated data and extract useful features therefrom. The first parameter vector (vector representation of humidity and temperature information) is placed in two-dimensional coordinates so that these complex high-dimensional data can be visualized, and patterns and trends therein can be easily identified. Further, in the two-dimensional coordinate system, the direction of change of these parameter vectors, whether the gradient rises or falls, will be identified by node analysis. The gradient rise or fall describes the trend and speed of the parameter vector over time (or other variable). If the gradient rises, this parameter is increasing; if the gradient decreases, this indicates that this parameter is decreasing. Such information is critical to understanding and predicting the change in state of the food product therein.
S4, obtaining second parameter information input by the sensing unit, wherein the second parameter information comprises but is not limited to volume information, route information, time information and heat information;
In S4, the system again obtains an additional set of parameter information from the sensor, specifically, the first parameter information is used to identify the "environment", and the second parameter information is the corresponding parameter of "after food is added to the environment". The second parameter information includes, but is not limited to: volume information, route information, time information, and thermal information. Volume information: this may refer to the size of the food package, affecting how the package is stored and transported, and thus affecting the time and cost of delivery. Route information: this includes the total distance of delivery, the projected route, etc., which is also necessary to calculate the expected delivery time and delivery schedule. Time information: this may refer to shipping time or estimated delivery time, etc., where time factors may be important in determining the quality of the food product and the efficiency of delivery. Heat information: during the distribution of cold chain food, heat is an important factor, as excessive heat may lead to reduced product quality and therefore is monitored and controlled in real time. After this series of information is obtained, the system will process and analyze it so that it can be effectively applied to the predictive model.
S5, vectorizing second parameter information by using an input gate of a preset long and short time sequence prediction unit, inputting the vectorized second parameter vector into a node analysis network, calibrating a first parameter vector and a second parameter vector in the node analysis network by using the time information, generating a corresponding travel progress vector by using distance data in the route information, loading the travel progress vector into the node analysis network, identifying vector offset of the travel progress vector and the second parameter vector on the node by using a cosine prediction algorithm, correspondingly adjusting node information of the travel progress vector on the node analysis network by using the vector offset, and generating a matched prediction result based on the node information.
In S5, vectorization of parameters: firstly, a preset long and short time sequence prediction unit is adopted to vectorize the input second parameter information. This means that the volume information, route information, time information and heat information just acquired are processed and converted into a form of vectors for further mathematical calculation or machine learning operations. After vectorizing these parameters we get a second parameter vector. Parameter calibration and travel progress vector generation: after having the corresponding second parameter vector, it is input into the node analysis network. In the node analysis network, the first parameter vector (vector of humidity and temperature information) and the second parameter vector are calibrated by time information, namely, the two types of data are ordered or synchronized according to time sequence, so as to form a series of data sets expressing detailed distribution process. Meanwhile, a corresponding journey progress vector is generated through route information (comprising distance information and the like in detail), and the journey progress vector is loaded into a node analysis network. The trip schedule vector is a vector representing the current delivery schedule by which the network can know the current state of the delivery schedule. Cosine prediction algorithm and generation of prediction results: after the steps are completed, the system adopts a cosine prediction algorithm to identify the vector offset of the travel progress vector and the second parameter vector on the node. How does the offset be understood? The vector offset actually describes the degree of variation between the two vectors, that is, this offset may indicate the gap between the actual progress of the delivery and the predicted progress. According to the offset, the system can properly adjust the node information of the travel progress vector on the node analysis network so as to enable the prediction model to be more in line with the actual distribution process. Based on the adjusted node information, the system finally generates a matched prediction result. This result will provide a more accurate prediction of cold chain food delivery, thereby helping businesses optimize delivery processes and improve efficiency.
In one embodiment, the step of performing merging calculation on the first parameter information based on a preset merging and sorting unit and performing vectorization on the first parameter information after merging calculation by combining a vector machine arranged in the merging and sorting unit to obtain a first parameter vector includes:
carrying out numerical processing on the first parameter information, and carrying out data collection on the numerical first parameter information through a merging and sorting unit to obtain parameter information with a digital group based on a time sequence collection;
carrying out vector conversion on first parameter information carrying a digital group by using a numerical value vector conversion method preset in a vector machine so as to generate a first parameter vector;
the method for digitally processing the first parameter information comprises the following steps:
performing information calibration on the humidity information and the temperature information of the first parameter information based on a time stamp;
and combining the values in the calibrated humidity information and the calibrated temperature information based on the time stamp to obtain a parameter information integrated with a digital group.
In the specific implementation process, the numerical treatment is carried out: first, it is necessary to digitize the first parameter information. Here, the first parameter information refers to temperature and humidity information. Before the digitizing process, the information needs to be calibrated with the corresponding time stamp, and the aim of the calibration is to correspond each piece of information with the specific time for generating the information, so that a set of information sets with time sequences is constructed. Then, by combining the values in the calibrated humidity information and temperature information, we obtain a piece of parameter information containing the digital set. Using a merge ordering unit: the processed first parameter information is transmitted to a preset merging and sorting unit. The merging and sorting function is to sort and integrate the first parameter information in numerical value to form a time sequence set, wherein the time sequence set contains parameter information of the digital group. Vectorization processing: then, the ordered parameter information is sent to a numerical vector conversion method preset in a vector machine for processing. Vectorization is a process of converting a specific value or array of values into a vector that reduces the dimensions of parameter information into a set of high-dimensional vectors that can be processed by a mathematical model. The method can improve the efficiency of information processing and lay a foundation for the following parameter prediction and analysis. Through the steps, a group of humidity and temperature information based on time sequence is integrated, ordered and converted into vectors, and then a first parameter vector is obtained. This vector can be used in subsequent data analysis and prediction.
In one embodiment, the step of performing vector conversion on the first parameter information carrying the digital group by using a numerical vector conversion method preset in a vector machine to generate a first parameter vector includes:
identifying corresponding humidity numbers and temperature numbers in a plurality of digital groups based on a quality quantization line, wherein the humidity numbers and the temperature numbers are numerical information of humidity information and temperature information respectively;
and vectorizing the quality quantization lines carrying humidity numbers and temperature numbers respectively and performing reverse processing to obtain a first parameter vector preset as gradient descent after reverse processing.
In a specific implementation process, the quality is identified: first, the parameter information is subjected to quality quantization processing. This step is based on a quality quantization line that can help carry the corresponding humidity and temperature numbers in the identification number set. The purpose of this step is to extract the temperature and humidity information and normalize it to numerical information, ready for vectorization. Vectorization and inverse processing: then, we further carry out vectorization processing on the humidity number and the temperature number on the quality quantification line, and the processing process comprises two steps of vector conversion and reverse processing. Vectorization is the conversion of each pair of humidity and temperature information into one vector in a multidimensional space for use in high-dimensional data operations. The inverse process refers to taking the inverse of the vector, i.e. taking its negative value, which is often useful in solving optimization problems, e.g. in gradient descent methods, we need to find the minimum point of the loss function, where if the function is a convex function, the vector is inverted, i.e. pointing in the direction of gradient descent. Finally, we obtain a first parameter vector preset as gradient descent through the above steps. This represents that the humidity and temperature information in the vector is processed into a form that can be used by machine learning models and the like, thereby laying the foundation for the following parameter prediction and analysis.
In one embodiment, the step of placing the first parameter vector in a two-dimensional coordinate for node analysis by using a preset node analysis network to identify the first parameter vector as gradient up or gradient down in the two-dimensional coordinate includes:
identifying a current time stamp in real time, and correspondingly updating an abscissa in a preset two-dimensional coordinate based on the current time stamp;
adding the first parameter vector on a two-dimensional coordinate based on a current time stamp, generating a plurality of nodes through refrigerating property preset on an ordinate, and correspondingly marking the numerical value of the first parameter vector based on the numerical values of the plurality of nodes so as to set the first parameter vector in a node analysis network;
carrying out refrigeration numerical analysis on the marked node numerical values and the first parameter vector numerical values to judge the numerical values of the first parameter vector and a preset suitable refrigeration threshold value so as to judge whether the first parameter vector accords with refrigeration or does not accord with refrigeration;
if the first parameter vector accords with refrigeration, the first parameter vector is in gradient ascending in the node analysis network, and otherwise, the first parameter vector is in gradient descending.
In a specific implementation, the goal of this step is to analyze the first parameter vector in two-dimensional coordinates using a predefined node analysis network to identify whether it is rising or falling in gradient. The following is a detailed explanation of the operational steps: updating the abscissa of the two-dimensional coordinates: first, the system will recognize the current timestamp in real time. Based on this time stamp, the abscissa in the two-dimensional coordinates is updated. In this coordinate system, the abscissa may represent time information. Placing a first parameter vector on two-dimensional coordinates: next, a first parameter vector is added to the two-dimensional coordinates based on the current timestamp, and a plurality of nodes are generated through preset refrigerating property. The nodes can place this information in the node analysis network against the values of the first parameter vector. And (3) carrying out cold storage numerical analysis: then, a refrigerated numerical analysis is performed based on the values of the nodes and the first parameter vector values. That is, the value of the first parameter vector is compared with a preset threshold value suitable for refrigeration, so as to determine whether the first parameter vector meets the refrigeration condition. If so, this temperature and humidity may be acceptable in cold chain transportation. Judging gradient trend: finally, if the first parameter vector is judged to be in accordance with the refrigeration condition, the first parameter vector is regarded as gradient rise in the node analysis network; otherwise, it is considered as a gradient descent. The gradient rise or fall here represents a potential for temperature or humidity changes during cold chain transport, which may affect the quality and preservation of the food product. Through the steps, the system can reasonably adjust the environment of the cold chain warehouse according to the actual first parameter vector (temperature and humidity information) so as to ensure the quality and safety of food.
The specific cold chain food distribution prediction method comprises the following steps: the refrigerating performance comprises four items of high-quality refrigerating conditions, common refrigerating conditions, inferior refrigerating conditions and non-conforming refrigerating conditions, the four items are correspondingly defined by preset numerical values, each refrigerating performance belongs to a numerical value interval, and the step of generating a plurality of nodes through the refrigerating performance preset by the ordinate comprises the following steps:
identifying a humidity number and a temperature number carried by a first parameter vector, and judging a corresponding interval of four refrigeration items in which the humidity number and the temperature number are located;
the first parameter vector is added with a plurality of sets of humidity numbers and temperature numbers according to the current time stamp along with the change of the current time stamp, and the humidity numbers and the temperature numbers of the sets are identified in real time and are changed on a coordinate system.
In particular, this section mentions four different refrigeration conditions, namely, a good refrigeration condition, a normal refrigeration condition, a bad refrigeration condition, and a condition that does not meet refrigeration. All four conditions have corresponding preset value intervals, and the set of definitions can be used for identifying and classifying various different refrigeration environments. The step of generating a cold storage property related node in a node analysis network comprises: and (3) identifying parameter vectors: first, the system identifies the humidity number and the temperature number carried by the first parameter vector. The two data points will be judged as to which preset value interval of cold storage property they are respectively located. Treatment over time: as the current timestamp changes, the first parameter vector will increment more sets of humidity and temperature figures according to this current timestamp. The system will recognize these new humidity and temperature numbers in real time and make corresponding changes in the two-dimensional coordinate system. In this way, real-time tracking and handling of various changes in refrigeration conditions can be maintained.
In one embodiment, the step of vectorizing the second parameter information by using an input gate of the preset long-short time sequence prediction unit to input the second parameter vector generated after vectorization to the node analysis network includes:
extracting characteristic information in the second parameter information by using an input gate of the long and short time sequence prediction unit, wherein the characteristic information comprises digital information corresponding to volume, route, time and heat;
carrying out partition vectorization on digital information corresponding to the volume, the route, the time and the heat after feature extraction to generate a plurality of parameter factors, and generating a time vector based on the plurality of parameter factors;
the time vectors are all loaded into the node analysis network as second parameter vectors, and the time vectors carry volume factors, route factors, time factors and heat factors.
In a specific implementation, the objective of this step is to vectorize the second parameter information using an input gate in a predetermined long-short timing prediction unit (Long Short Term Memory, LSTM). And then, inputting the second parameter vector generated after vectorization into a node analysis network. This process may be embodied as the following steps: feature extraction: first, feature information in the second parameter information is extracted using an input gate of the LSTM. These characteristic information include digital information corresponding to volume, route, time and heat. These are important factors that can affect the environment and efficiency of the cold chain transportation process. Partition vectorization: and then, converting the extracted volume, route, time and heat information into corresponding digital information, and carrying out partition vectorization. Partition vectorization converts this digital information into a set of parameter factors. Based on these parameter factors, we can generate a time vector. Input node analysis network: and finally, loading the generated time vector as a second parameter vector into the node analysis network. The time vector carries volume factors, route factors, time factors and heat factors which are important parameters for determining the refrigerating condition. By the method, the second parameter information (such as volume, route, time, heat and the like) can be taken into consideration of the node analysis network, so that the refrigeration environment and the change trend thereof can be analyzed more comprehensively and accurately.
In one embodiment, the step of calibrating the first parameter vector and the second parameter vector in the node analysis network by the time information includes:
determining a starting node in a node analysis network based on a time factor in the second parameter vector;
generating a period of prediction time based on the time factor and the route factor in the second parameter vector on the premise that the abscissa is time;
determining the refrigeration of the ordinate of the matched node analysis network based on the volume factor and the heat factor;
and determining cold storage property through the starting node, the predicted time period and the volume factor and the heat factor to generate a second parameter vector in the node analysis network.
In a specific implementation, a starting node is determined: first, a starting node in the node analysis network is determined using the time factor in the second parameter vector. This starting node is the starting point of the analysis and is derived based on the value of the time factor. Generating a predicted time period: then, based on the time factor and the route factor in the second parameter vector, a predicted time period is generated on the premise of representing time on the abscissa. This period is used for subsequent parameter analysis and prediction. Determine the refrigeration properties on the ordinate: then, based on the volume factor and the thermal factor in the second parameter vector, the refrigeration of the ordinate of the matched node analysis network is determined. This means that the refrigeration conditions (good, normal, inferior, not in line with refrigeration) represented by the ordinate are dynamically adjusted in terms of volume and heat. Generating a second parameter vector: finally, a second parameter vector in the node analysis network is generated by the starting node, the predicted time period and the determined cold storage property. This parameter vector contains information about volume, course, time and heat after conversion, which is key input data used to further analyze the refrigeration conditions.
In one embodiment, the step of generating a corresponding travel progress vector according to the distance data in the route information and loading the travel progress vector into the node analysis network includes:
a corresponding distance schedule is called through the distance data, so that travel schedule data corresponding to the travel schedule vector are correspondingly obtained from the distance schedule, the travel schedule vector is obtained after the travel schedule data are vectorized, and the travel schedule vector is used as a preset distribution standard;
and loading the travel progress vector into a node analysis network, wherein the starting node of the travel progress vector is consistent with the starting node of the second parameter vector.
In a specific implementation process, a travel table is called and a progress vector is generated: first, the distance data in the route information is used to retrieve the corresponding distance schedule. Then, the corresponding travel schedule data is extracted from the distance travel table. And vectorizing the travel progress data to obtain a travel progress vector. This trip schedule vector may be used as a preset delivery criteria. Load node analysis network: and loading the travel progress vector generated in the last step into a node analysis network. The start node of the travel progress vector coincides with the start node of the previously generated second parameter vector. Through these steps, the trip schedule is effectively quantified and introduced into the node analysis network so as to fully consider and cope with the impact of the trip distance and schedule on the refrigeration condition of the goods during the cold chain transportation.
In one embodiment, the step of identifying a vector offset between the travel progress vector and the second parameter vector on the node by using a cosine prediction algorithm, correspondingly adjusting node information of the travel progress vector on the node analysis network by using the vector offset, and generating a matched prediction result based on the node information includes:
identifying an included angle between the travel progress vector and the second parameter vector in the node analysis network, wherein the included angle is only confirmed from the vector line of the travel progress vector and the second parameter vector, and the obtained included angle is a vector offset;
identifying each offset node generating a vector offset, and calculating the offset node and the vector offset by a cosine prediction algorithm to calculate the cosine development direction of the offset node;
based on the cosine development direction and combined with the refrigerating property and the predicted time period of the second parameter vector, correspondingly generating terminal node data of the predicted time period, wherein the terminal node data comprises a delivery time and a refrigerating property value;
and correspondingly adjusting the travel progress vector based on the final node data, and displaying the travel progress vector adjusted in real time as a prediction result.
In a specific process, the offset is identified: first, an angle between the travel progress vector and the second parameter vector is identified in the node analysis network. This angle is measured from the vector line of the travel progress vector and the second parameter vector, which is a measurement called the vector offset. Cosine prediction algorithm calculation: each offset node that produces a vector offset is identified and these offset nodes and vector offsets are applied to a cosine prediction algorithm for calculation. The purpose of this calculation is to derive the cosine development direction of the offset node. Generating end node data of a predicted time period: and generating terminal node data of a predicted time period according to the calculated cosine development direction and combining the refrigerating property of the second parameter vector and the predicted time period. The data for this end node includes the predicted delivery time and the refrigerated value. Adjusting a travel progress vector: and finally, correspondingly adjusting the travel progress vector according to the generated terminal node data. And displaying the adjusted travel progress vector in real time as a prediction result. Through fine control and adjustment of the travel progress vector, accurate prediction and planning of the cold chain transportation process are facilitated, and therefore effective improvement and optimization of refrigeration quality are achieved.
Referring to fig. 2, a block diagram of a cold chain food delivery prediction system according to the present invention is provided, where the system includes:
a first acquiring unit 1 for acquiring first parameter information input by the sensing unit, the first parameter information including but not limited to humidity information and temperature information;
the merging unit 2 is used for merging and calculating the first parameter information based on a preset merging and sorting unit, and vectorizing the first parameter information after merging and calculating by combining a vector machine arranged in the merging and sorting unit to obtain a first parameter vector;
the node network unit 3 is configured to place the first parameter vector in a two-dimensional coordinate by using a preset node analysis network to perform node analysis, so as to identify that the first parameter vector is gradient rising or gradient falling in the two-dimensional coordinate;
a second acquiring unit 4 for acquiring second parameter information input again by the sensing unit, the second parameter information including but not limited to volume information, route information, time information and heat information;
the model processing unit 5 is configured to perform vectorization on the second parameter information by using an input gate of the preset long and short time sequence prediction unit, so as to input the second parameter vector generated after vectorization to the node analysis network, calibrate the first parameter vector and the second parameter vector in the node analysis network through the time information, generate a corresponding travel progress vector through distance data in the route information, load the travel progress vector into the node analysis network, identify a vector offset of the travel progress vector and the second parameter vector on the node by adopting a cosine prediction algorithm, correspondingly adjust node information of the travel progress vector on the node analysis network through the vector offset, and generate a matched prediction result based on the node information.
In summary, the first parameter information input by the sensing unit is obtained, where the first parameter information includes, but is not limited to, humidity information and temperature information; the method comprises the steps of carrying out merging calculation on first parameter information based on a preset merging and sorting unit, and carrying out vectorization on the first parameter information after merging calculation by combining a vector machine arranged in the merging and sorting unit to obtain a first parameter vector; the first parameter vector is placed in a two-dimensional coordinate by adopting a preset node analysis network to carry out node analysis, so that the first parameter vector is identified as gradient rising or gradient falling in the two-dimensional coordinate; acquiring second parameter information input by the sensing unit, wherein the second parameter information comprises, but is not limited to, volume information, route information, time information and heat information; vectorizing second parameter information by using an input gate of a preset long and short time sequence prediction unit, inputting the second parameter vector generated after vectorization into a node analysis network, calibrating a first parameter vector and a second parameter vector in the node analysis network by using the time information, generating a corresponding travel progress vector by using distance data in the route information, loading the travel progress vector into the node analysis network, identifying vector offset of the travel progress vector and the second parameter vector on the node by using a cosine prediction algorithm, correspondingly adjusting node information of the travel progress vector on the node analysis network by using the vector offset, and generating a matched prediction result based on the node information; by accurately processing and analyzing parameters such as humidity, temperature, volume, route, time, heat and the like, optimal control of delivery time and path is realized, delivery precision and efficiency are improved, and food loss and delivery cost are reduced.
The foregoing description is only of the preferred embodiments of the present invention and is not intended to limit the scope of the invention, and all equivalent structures or equivalent processes using the descriptions and drawings of the present invention or direct or indirect application in other related technical fields are included in the scope of the present invention.
Claims (6)
1. A method for predicting the delivery of a cold chain food product, comprising the steps of:
acquiring first parameter information input by a sensing unit, wherein the first parameter information comprises, but is not limited to, humidity information and temperature information;
the method comprises the steps of carrying out merging calculation on first parameter information based on a preset merging and sorting unit, and carrying out vectorization on the first parameter information after merging calculation by combining a vector machine arranged in the merging and sorting unit to obtain a first parameter vector;
the first parameter vector is placed in a two-dimensional coordinate by adopting a preset node analysis network to carry out node analysis, so that the first parameter vector is identified as gradient rising or gradient falling in the two-dimensional coordinate;
acquiring second parameter information input by the sensing unit, wherein the second parameter information comprises, but is not limited to, volume information, route information, time information and heat information;
Vectorizing second parameter information by using an input gate of a preset long and short time sequence prediction unit, inputting the second parameter vector generated after vectorization into a node analysis network, calibrating a first parameter vector and a second parameter vector in the node analysis network by using the time information, generating a corresponding travel progress vector by using distance data in the route information, loading the travel progress vector into the node analysis network, identifying vector offset of the travel progress vector and the second parameter vector on the node by using a cosine prediction algorithm, correspondingly adjusting node information of the travel progress vector on the node analysis network by using the vector offset, and generating a matched prediction result based on the node information;
the method comprises the steps of carrying out merging calculation on first parameter information based on a preset merging and sorting unit, carrying out vectorization on the first parameter information after merging calculation by combining a vector machine arranged in the merging and sorting unit, and obtaining a first parameter vector, and comprises the following steps:
carrying out numerical processing on the first parameter information, and carrying out data collection on the numerical first parameter information through a merging and sorting unit to obtain parameter information with a digital group based on a time sequence collection;
Carrying out vector conversion on first parameter information carrying a digital group by using a numerical value vector conversion method preset in a vector machine so as to generate a first parameter vector;
the method for digitally processing the first parameter information comprises the following steps:
performing information calibration on the humidity information and the temperature information of the first parameter information based on a time stamp;
combining the values in the calibrated humidity information and the calibrated temperature information based on the time stamp to obtain parameter information integrated with a digital group;
the step of placing the first parameter vector in a two-dimensional coordinate by adopting a preset node analysis network to perform node analysis so as to identify the first parameter vector as gradient rising or gradient falling in the two-dimensional coordinate comprises the following steps:
identifying a current time stamp in real time, and correspondingly updating an abscissa in a preset two-dimensional coordinate based on the current time stamp;
adding the first parameter vector on a two-dimensional coordinate based on a current time stamp, generating a plurality of nodes through refrigerating property preset on an ordinate, and correspondingly marking the numerical value of the first parameter vector based on the numerical values of the plurality of nodes so as to set the first parameter vector in a node analysis network;
Carrying out refrigeration numerical analysis on the marked node numerical values and the first parameter vector numerical values to judge the numerical values of the first parameter vector and a preset suitable refrigeration threshold value so as to judge whether the first parameter vector accords with refrigeration or does not accord with refrigeration;
if the first parameter vector accords with refrigeration, the first parameter vector is in gradient ascending in the node analysis network, otherwise, the first parameter vector is in gradient descending;
wherein: the refrigerating performance comprises four items of high-quality refrigerating conditions, common refrigerating conditions, inferior refrigerating conditions and non-conforming refrigerating conditions, the four items are correspondingly defined by preset numerical values, each refrigerating performance belongs to a numerical value interval, and the step of generating a plurality of nodes through the refrigerating performance preset by the ordinate comprises the following steps:
identifying a humidity number and a temperature number carried by a first parameter vector, and judging a corresponding interval of four refrigeration items in which the humidity number and the temperature number are located;
the first parameter vector can be added with a plurality of sets of humidity numbers and temperature numbers according to the current timestamp along with the change of the current timestamp, and the humidity numbers and the temperature numbers of the sets are identified in real time and are changed on a coordinate system;
Identifying vector offset of the travel progress vector and the second parameter vector on the node by adopting a cosine prediction algorithm, correspondingly adjusting node information of the travel progress vector on a node analysis network through the vector offset, and generating a matched prediction result based on the node information, wherein the method comprises the steps of:
identifying an included angle between the travel progress vector and the second parameter vector in the node analysis network, wherein the included angle is only confirmed from the vector line of the travel progress vector and the second parameter vector, and the obtained included angle is a vector offset;
identifying each offset node generating a vector offset, and calculating the offset node and the vector offset by a cosine prediction algorithm to calculate the cosine development direction of the offset node;
based on the cosine development direction and combined with the refrigerating property and the predicted time period of the second parameter vector, correspondingly generating terminal node data of the predicted time period, wherein the terminal node data comprises a delivery time and a refrigerating property value;
and correspondingly adjusting the travel progress vector based on the final node data, and displaying the travel progress vector adjusted in real time as a prediction result.
2. The method of claim 1, wherein the step of vector-converting the first parameter information carrying the digital group by using a numerical vector conversion method preset in a vector machine to generate the first parameter vector comprises:
Identifying corresponding humidity numbers and temperature numbers in a plurality of digital groups based on a quality quantization line, wherein the humidity numbers and the temperature numbers are numerical information of humidity information and temperature information respectively;
and vectorizing the quality quantization lines carrying humidity numbers and temperature numbers respectively and performing reverse processing to obtain a first parameter vector preset as gradient descent after reverse processing.
3. The method for predicting cold chain food distribution according to claim 1, wherein the step of vectorizing the second parameter information by using an input gate of a preset long-short time sequence prediction unit to input the vectorized second parameter vector to the node analysis network comprises:
extracting characteristic information in the second parameter information by using an input gate of the long and short time sequence prediction unit, wherein the characteristic information comprises digital information corresponding to volume, route, time and heat;
carrying out partition vectorization on digital information corresponding to the volume, the route, the time and the heat after feature extraction to generate a plurality of parameter factors, and generating a time vector based on the plurality of parameter factors;
the time vectors are all loaded into the node analysis network as second parameter vectors, and the time vectors carry volume factors, route factors, time factors and heat factors.
4. A cold chain food distribution prediction method according to claim 3, characterized in that the step of calibrating the first and second parameter vectors in the node analysis network by means of the time information comprises:
determining a starting node in a node analysis network based on a time factor in the second parameter vector;
generating a period of prediction time based on the time factor and the route factor in the second parameter vector on the premise that the abscissa is time;
determining the refrigeration of the ordinate of the matched node analysis network based on the volume factor and the heat factor;
and determining cold storage property through the starting node, the predicted time period and the volume factor and the heat factor to generate a second parameter vector in the node analysis network.
5. The cold chain food distribution prediction method according to claim 4, wherein the step of generating a corresponding trip progress vector from the distance data in the route information and loading the trip progress vector in the node analysis network comprises:
a corresponding distance schedule is called through the distance data, so that travel schedule data corresponding to the travel schedule vector are correspondingly obtained from the distance schedule, the travel schedule vector is obtained after the travel schedule data are vectorized, and the travel schedule vector is used as a preset distribution standard;
And loading the travel progress vector into a node analysis network, wherein the starting node of the travel progress vector is consistent with the starting node of the second parameter vector.
6. A cold chain food distribution prediction system, comprising:
a first acquisition unit for acquiring first parameter information input by the sensing unit, wherein the first parameter information comprises, but is not limited to, humidity information and temperature information;
the merging unit is used for merging and calculating the first parameter information based on a preset merging and sorting unit, and vectorizing the first parameter information after merging and calculating by combining a vector machine arranged in the merging and sorting unit to obtain a first parameter vector;
the node network unit is used for placing the first parameter vector in a two-dimensional coordinate by adopting a preset node analysis network to perform node analysis so as to identify that the first parameter vector is in gradient ascending or gradient descending in the two-dimensional coordinate;
a second acquisition unit for acquiring second parameter information input by the sensing unit, wherein the second parameter information comprises, but is not limited to, volume information, route information, time information and heat information;
the model processing unit is used for vectorizing second parameter information by utilizing an input gate of the preset long and short time sequence prediction unit, inputting the second parameter vector generated after vectorizing into the node analysis network, calibrating the first parameter vector and the second parameter vector in the node analysis network by the time information, generating a corresponding travel progress vector by distance data in the route information, loading the corresponding travel progress vector into the node analysis network, identifying vector offset of the travel progress vector and the second parameter vector on the node by adopting a cosine prediction algorithm, correspondingly adjusting node information of the travel progress vector on the node analysis network by the vector offset, and generating a matched prediction result based on the node information;
The method comprises the steps of carrying out merging calculation on first parameter information based on a preset merging and sorting unit, carrying out vectorization on the first parameter information after merging calculation by combining a vector machine arranged in the merging and sorting unit, and obtaining a first parameter vector, and comprises the following steps:
carrying out numerical processing on the first parameter information, and carrying out data collection on the numerical first parameter information through a merging and sorting unit to obtain parameter information with a digital group based on a time sequence collection;
carrying out vector conversion on first parameter information carrying a digital group by using a numerical value vector conversion method preset in a vector machine so as to generate a first parameter vector;
the method for digitally processing the first parameter information comprises the following steps:
performing information calibration on the humidity information and the temperature information of the first parameter information based on a time stamp;
combining the values in the calibrated humidity information and the calibrated temperature information based on the time stamp to obtain parameter information integrated with a digital group;
the step of placing the first parameter vector in a two-dimensional coordinate by adopting a preset node analysis network to perform node analysis so as to identify the first parameter vector as gradient rising or gradient falling in the two-dimensional coordinate comprises the following steps:
Identifying a current time stamp in real time, and correspondingly updating an abscissa in a preset two-dimensional coordinate based on the current time stamp;
adding the first parameter vector on a two-dimensional coordinate based on a current time stamp, generating a plurality of nodes through refrigerating property preset on an ordinate, and correspondingly marking the numerical value of the first parameter vector based on the numerical values of the plurality of nodes so as to set the first parameter vector in a node analysis network;
carrying out refrigeration numerical analysis on the marked node numerical values and the first parameter vector numerical values to judge the numerical values of the first parameter vector and a preset suitable refrigeration threshold value so as to judge whether the first parameter vector accords with refrigeration or does not accord with refrigeration;
if the first parameter vector accords with refrigeration, the first parameter vector is in gradient ascending in the node analysis network, otherwise, the first parameter vector is in gradient descending;
wherein: the refrigerating performance comprises four items of high-quality refrigerating conditions, common refrigerating conditions, inferior refrigerating conditions and non-conforming refrigerating conditions, the four items are correspondingly defined by preset numerical values, each refrigerating performance belongs to a numerical value interval, and the step of generating a plurality of nodes through the refrigerating performance preset by the ordinate comprises the following steps:
identifying a humidity number and a temperature number carried by a first parameter vector, and judging a corresponding interval of four refrigeration items in which the humidity number and the temperature number are located;
The first parameter vector can be added with a plurality of sets of humidity numbers and temperature numbers according to the current timestamp along with the change of the current timestamp, and the humidity numbers and the temperature numbers of the sets are identified in real time and are changed on a coordinate system;
identifying vector offset of the travel progress vector and the second parameter vector on the node by adopting a cosine prediction algorithm, correspondingly adjusting node information of the travel progress vector on a node analysis network through the vector offset, and generating a matched prediction result based on the node information, wherein the method comprises the steps of:
identifying an included angle between the travel progress vector and the second parameter vector in the node analysis network, wherein the included angle is only confirmed from the vector line of the travel progress vector and the second parameter vector, and the obtained included angle is a vector offset;
identifying each offset node generating a vector offset, and calculating the offset node and the vector offset by a cosine prediction algorithm to calculate the cosine development direction of the offset node;
based on the cosine development direction and combined with the refrigerating property and the predicted time period of the second parameter vector, correspondingly generating terminal node data of the predicted time period, wherein the terminal node data comprises a delivery time and a refrigerating property value;
And correspondingly adjusting the travel progress vector based on the final node data, and displaying the travel progress vector adjusted in real time as a prediction result.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310953977.4A CN116663753B (en) | 2023-08-01 | 2023-08-01 | Cold chain food distribution prediction method and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310953977.4A CN116663753B (en) | 2023-08-01 | 2023-08-01 | Cold chain food distribution prediction method and system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN116663753A CN116663753A (en) | 2023-08-29 |
CN116663753B true CN116663753B (en) | 2023-10-20 |
Family
ID=87721061
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310953977.4A Active CN116663753B (en) | 2023-08-01 | 2023-08-01 | Cold chain food distribution prediction method and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116663753B (en) |
Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108846498A (en) * | 2018-03-30 | 2018-11-20 | 东南大学 | Drug traceability system and method based on GIS Yu temperature and humidity integration technology |
CN110110909A (en) * | 2019-04-23 | 2019-08-09 | 中山大学 | Cold chain product quality forecast and intelligent distribution method based on RFID and extreme learning machine |
CN110135783A (en) * | 2019-05-15 | 2019-08-16 | 贵州西部农产品交易中心有限公司 | A kind of tea industry intelligence Cold Chain Logistics information service platform |
WO2020164740A1 (en) * | 2019-02-15 | 2020-08-20 | Huawei Technologies Co., Ltd. | Methods and systems for automatically selecting a model for time series prediction of a data stream |
CN112766561A (en) * | 2021-01-15 | 2021-05-07 | 东南大学 | Generating type confrontation track prediction method based on attention mechanism |
CN112990595A (en) * | 2021-03-30 | 2021-06-18 | 北京嘀嘀无限科技发展有限公司 | Travel time prediction method, travel time prediction device, storage medium and electronic equipment |
CN113435593A (en) * | 2021-06-29 | 2021-09-24 | 广州博通信息技术有限公司 | Refrigeration equipment frosting prediction method based on sensor time sequence data analysis |
KR102327747B1 (en) * | 2021-07-30 | 2021-11-17 | 이석용 | System and Method for managing shelf life and inventory of food distributed through cold chain based on demand prediction model using big data |
CN113807780A (en) * | 2021-09-04 | 2021-12-17 | 刘伟 | Cold chain transportation method in big data environment and gas column bag for cold chain transportation |
WO2023087508A1 (en) * | 2021-11-18 | 2023-05-25 | 广州工商学院 | Cold chain logistics monitoring method and system, and computer device |
CN116167687A (en) * | 2023-04-25 | 2023-05-26 | 成都运荔枝科技有限公司 | Intelligent scheduling method for cold chain system |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9373099B2 (en) * | 2013-03-12 | 2016-06-21 | Tcp Reliable, Inc. | Network of participants in a shipment cold-chain |
-
2023
- 2023-08-01 CN CN202310953977.4A patent/CN116663753B/en active Active
Patent Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108846498A (en) * | 2018-03-30 | 2018-11-20 | 东南大学 | Drug traceability system and method based on GIS Yu temperature and humidity integration technology |
WO2020164740A1 (en) * | 2019-02-15 | 2020-08-20 | Huawei Technologies Co., Ltd. | Methods and systems for automatically selecting a model for time series prediction of a data stream |
CN110110909A (en) * | 2019-04-23 | 2019-08-09 | 中山大学 | Cold chain product quality forecast and intelligent distribution method based on RFID and extreme learning machine |
CN110135783A (en) * | 2019-05-15 | 2019-08-16 | 贵州西部农产品交易中心有限公司 | A kind of tea industry intelligence Cold Chain Logistics information service platform |
CN112766561A (en) * | 2021-01-15 | 2021-05-07 | 东南大学 | Generating type confrontation track prediction method based on attention mechanism |
CN112990595A (en) * | 2021-03-30 | 2021-06-18 | 北京嘀嘀无限科技发展有限公司 | Travel time prediction method, travel time prediction device, storage medium and electronic equipment |
CN113435593A (en) * | 2021-06-29 | 2021-09-24 | 广州博通信息技术有限公司 | Refrigeration equipment frosting prediction method based on sensor time sequence data analysis |
KR102327747B1 (en) * | 2021-07-30 | 2021-11-17 | 이석용 | System and Method for managing shelf life and inventory of food distributed through cold chain based on demand prediction model using big data |
CN113807780A (en) * | 2021-09-04 | 2021-12-17 | 刘伟 | Cold chain transportation method in big data environment and gas column bag for cold chain transportation |
WO2023087508A1 (en) * | 2021-11-18 | 2023-05-25 | 广州工商学院 | Cold chain logistics monitoring method and system, and computer device |
CN116167687A (en) * | 2023-04-25 | 2023-05-26 | 成都运荔枝科技有限公司 | Intelligent scheduling method for cold chain system |
Non-Patent Citations (3)
Title |
---|
Zhaoyuan Xu.Forecast of Cold Chain Logistics Demand Based on GM (1,1) Model.《2021 3rd International Academic Exchange Conference on Science and Technology Innovation (IAECST)》.2021,全文. * |
基于AHPSO-SVM的农产品冷链物流质量安全预警模型;杨玮;偶雅楠;岳婷;杨甜;李沁;;包装工程(第05期);全文 * |
基于物联网的医疗冷链监测应用探讨;丁忠;姚青岭;严子禾;鲁晓杰;孙炜一;;中国医疗设备(第10期);全文 * |
Also Published As
Publication number | Publication date |
---|---|
CN116663753A (en) | 2023-08-29 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN116383450B (en) | Railway and highway logistics transportation information comprehensive management system | |
JP2022500769A (en) | Power system heat load prediction method and prediction device | |
CN108241901B (en) | Transformer early warning evaluation method and device based on prediction data | |
WO2015136586A1 (en) | Factor analysis device, factor analysis method, and factor analysis program | |
CN110751339A (en) | Method and device for predicting corrosion rate of pipeline and computer equipment | |
CN115824993B (en) | Method and device for determining water body chemical oxygen demand, computer equipment and medium | |
US20210394247A1 (en) | Method and system for forming a stamped component using a stamping simulation model | |
CN112985494A (en) | Cold chain wisdom logistics transportation on-line real-time supervision cloud platform based on big data and artificial intelligence | |
CN114417954A (en) | Information processing method and system for improving food detection effect | |
CN116663753B (en) | Cold chain food distribution prediction method and system | |
CN116629726B (en) | Environment monitoring management method and system for precise copper pipe packaging workshop | |
US11580414B2 (en) | Factor analysis device, factor analysis method, and storage medium on which program is stored | |
CN117592656A (en) | Carbon footprint monitoring method and system based on carbon data accounting | |
CN107562778A (en) | A kind of outlier excavation method based on deviation feature | |
CN115935285A (en) | Multi-element time series anomaly detection method and system based on mask map neural network model | |
CN113312587B (en) | Sensor acquisition data missing value processing method based on ARIMA prediction and regression prediction | |
CN115345256B (en) | Industrial product testing system applied to intelligent manufacturing | |
Cho et al. | Quality prediction modeling of plastic extrusion process | |
Denisova et al. | Supervised multichannel image classification algorithm using hierarchical histogram representation | |
CN115752243A (en) | Measurement data fusion method | |
CN114676868A (en) | Logistics cargo quantity prediction method and device, computer equipment and storage medium | |
CN114422450A (en) | Network flow analysis method and device based on multi-source network flow data | |
JP7118690B2 (en) | Manufacturing method, information processing device, relational expression calculation device, and manufacturing system | |
WO2019142344A1 (en) | Analysis device, analysis method, and recording medium | |
CN117290799B (en) | Enterprise purchase management method and system based on big data |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |