CN115274131A

CN115274131A - Epidemic prevention optimization method for epidemic infectious disease of logistics system based on big data

Info

Publication number: CN115274131A
Application number: CN202210795479.7A
Authority: CN
Inventors: 王海灵; 杨亚宁
Original assignee: Xinjiang University
Current assignee: Xinjiang University
Priority date: 2022-07-06
Filing date: 2022-07-06
Publication date: 2022-11-01

Abstract

The invention relates to a logistics system epidemic disease epidemic prevention optimization method based on big data, which comprises the steps of data acquisition, data matching, data evaluation and risk level confirmation, spot inspection proportion and disinfection grade determination according to the confirmed risk level, and post-posting quarantine certification terminal after disinfection effect evaluation. The epidemic prevention optimization method provided by the invention adopts a combined model of random forest + KNN + expert grading method to predict risks, combines big data to quickly evaluate the epidemic disease risk level of the logistics system, improves the inspection and quarantine working speed of the entry logistics system, designs the authentication mark and the mutual authentication standard of the inspection and quarantine and disinfection effects of the logistics system, provides the safety level certification of the logistics system and five-element inspection, quarantine, disinfection and disinfection, prevents the entry logistics system from repeated inspection, quarantine, disinfection or excessive disinfection, scientifically and accurately improves the inspection efficiency of the logistics system, reduces the service cost, and improves the entry convenience and optimizes the service dual-cycle operator environment.

Description

Epidemic prevention optimization method for epidemic infectious disease of logistics system based on big data

Technical Field

The invention relates to the technical field of epidemic prevention, in particular to a logistics system epidemic prevention optimization method based on big data.

Background

The domestic quarantine work of the inbound personnel, the vehicles and the articles mainly comprises two stages of inspection reporting and quarantine, the information collection in the inspection reporting stage has a method with small data statistical specifications at present, and the inspection and quarantine process has standardized personnel allocation rules and work flows. In recent years, the epidemic prevention policy for epidemic infectious diseases related to logistics activities is improved, and the epidemic prevention and killing work of each function related to epidemic infectious diseases in logistics activities is specified. In addition, the established related flow of epidemic prevention of epidemic diseases of people, vehicles and goods is clearer for the related work flow of epidemic prevention of epidemic diseases of the inbound people and goods, the information collection content of the inbound people and goods is determined, and partial moving tracks of the logistics system can be found out.

However, the existing related method for epidemic prevention of epidemic infectious diseases of logistics systems in China still has the following defects: 1. the intelligent degree of a risk assessment mechanism of the personnel and goods to be detected in the inspection and quarantine work is not high, and the risk assessment capability of other elements in the logistics system is not enough. In recent years, customs has made more comprehensive and standardized information collection documents, but the extraction of the document information is still completed manually; when assessing the risk level, qualitative analysis is mainly used, and accurate assessment is difficult to achieve. 2. The goods sampling inspection proportion lacks a scientific and accurate mechanism, and different sampling inspection proportions are not adopted for goods from different risk level areas, so that the phenomenon of one-time cutting occurs. After the epidemic situation of the infectious disease occurs, since the selective inspection is feared to cause the missed inspection of the logistics system from the high-risk area, the transmission is caused, so that most logistics nodes adopt the full inspection, and a great deal of time and cost are consumed for the logistics system from the low-risk area, and the efficiency is low. 3. Killing mode and killing effect evaluation. On one hand, the conditions of the logistics system and the cross infection of epidemic infectious disease pathogens among all elements are not considered, the disinfection and killing of all elements of the logistics system are not unified, the virus propagation risk exists when all elements of the logistics system from a high-risk area are not disinfected, and a large amount of public resources are consumed when all elements of the logistics system from a low-risk area are disinfected. On the other hand, the existing disinfection and killing effect evaluation is long in time consumption, disinfection and killing effect evaluation specified in various domestic sanitary regulations is carried out in a laboratory at present, the number of pathogens in a sample before and after target pathogen disinfection needs to be accurately detected, if the disinfection and killing effect evaluation is not carried out, the disinfection and killing effect cannot be guaranteed, and if the existing evaluation method is adopted, cargo overstock, customs area congestion, entry logistics unsmooth, efficiency is low and the environment of a carrier is poor due to long time. In order to solve the problem that the time for detecting the killing effect is too long, a student uses poliovirus as a unified standard control sample to judge the killing effect, the poliovirus and the target pathogen kill simultaneously, and if the killing effect of the killed poliovirus meets the sanitary requirement, the target pathogen killing effect is determined to be qualified; the students propose that the adenovirus marked by GFP can play the same role as poliomyelitis, so that the experimental result is easy to obtain and the time is short; the detection of poliovirus still needs to be carried out in a laboratory for a long time at present. 4. The feedback mechanism in the quarantine work is not perfect, and the problems of the quarantined personnel, goods and vehicles in the later circulation link are not used as the feedback information of the quarantine work, so that the quarantine work policy is continuously adjusted to improve the quarantine safety level. The historical data of epidemic infectious disease pathogen detection results of people, goods and vehicles entering through the entry logistics system and the data of public health crisis times caused by entry of related logistics system elements are basically not used as the basis for formulating the entry quarantine reference standard, so that dynamic adjustment and iterative optimization of a quarantine mechanism are hardly performed. The scientific accuracy and the safety of quarantine work are lower. 5. Because there is no mutual authentication standard for disinfection, evaluation and authentication, administrative administration responsibility is divided and is uncoordinated. Because the mutual authentication standards of disinfection evaluation authentication and authentication are not available, the trust degree of quarantine work among next-level administrative organizations in various regions and regions in China is not enough, so that logistics nodes arranged in the regions where the inbound logistics passes are required to be quarantined and disinfected, repeated quarantine and disinfection or excessive disinfection occurs, public resources are wasted, and the efficiency of a logistics system is reduced.

Disclosure of Invention

The invention aims to provide a logistics system epidemic prevention optimization method based on big data, which can solve the technical problems mentioned in the background technology.

In order to achieve the purpose, the invention provides the following technical scheme, and the logistics system epidemic prevention optimization method based on big data comprises the following steps:

data acquisition: collecting epidemic risk data of elements of a logistics system;

data matching: matching the epidemic risk data with an epidemic risk evaluation database to obtain the risk level of elements of the logistics system;

and (3) evaluating data: confirming the overall risk level of the logistics system by adopting an expert judgment method and a KNN algorithm combined model;

matching and detecting the sampling inspection proportion of the elements of the logistics system according to the overall risk level;

matching killing grades according to the overall risk grade, and killing;

evaluating the killing effect;

and storing the data for retrieval and posting a quarantine certification terminal.

Preferably, the elements of the logistics system comprise at least two of the following: driver, goods packing, turnover tool and vehicle.

Preferably, the epidemic disease risk assessment database is derived from epidemic disease risk condition information published by government departments in various regions.

As a preferred scheme, before data matching, the number of epidemic infectors in an accurate country is used as a training sample, a random forest regression algorithm is adopted to obtain a prediction model, and the prediction model is adopted to process a risk database of the epidemic infectious disease in a non-accurate country to obtain a risk value database of the epidemic infectious disease in the geographic position.

In the data evaluation, the scores of low risk, medium risk and high risk grade are obtained through an expert judgment method, the elements of the logistics system are assigned to obtain a risk matrix of the logistics system, and the risk matrix is input into a KNN algorithm to obtain the overall risk grade of the logistics system.

Preferably, the known disease source and the unknown disease source are matched with a killing method according to the confirmed risk grade, and killing is carried out.

As a preferred scheme, the evaluation of the killing effect adopts: before killing, spraying the aqueous solution containing GFP-labeled adenovirus onto the surface of elements of the logistics system, collecting surface samples, calculating the amount of adenovirus, after killing, collecting the surface samples of the objects again, and calculating the amount of adenovirus.

As a preferred scheme, the quarantine certification terminal comprises a wireless radio frequency information base, wherein the wireless radio frequency information base comprises a label making unit, making time, a record number, and risk information, risk grades, killing modes and killing effects of elements of each logistics system. The docket number is encrypted by a key.

Compared with the prior epidemic prevention method for epidemic infectious diseases of a logistics system, the invention has the beneficial effects that: aiming at the defects of the related control mode of the epidemic infectious disease of the current entry logistics system, the invention fully utilizes the available data which influences the logistics system to carry the epidemic infectious disease pathogen, considers the infection probability of the epidemic infectious disease pathogen among all elements in the foreign logistics vehicle, rapidly evaluates the whole logistics system and the risk level of the epidemic infectious disease of all the elements by using a random forest regression, an expert judgment method and a KNN combined algorithm, optimizes a killing method and a killing effect evaluation method of the logistics system, improves the quarantine work efficiency of an entry port, and designs a logistics system quarantine result to prove that a safety certificate is provided for the entry logistics system to prevent the entry logistics system from being repeatedly quarantined and killed.

The invention provides a random forest regression, an expert judgment method and a KNN combined algorithm, wherein the random forest provides a more perfect data source for the expert judgment method and the KNN, the expert judgment method is combined with the KNN qualitative and quantitative method, the risks and the overall risks of all elements of a logistics system are effectively quantified, different from the single random forest regression, the results obtained by the random forest regression quantitative analysis are qualitatively processed by the expert judgment method, the judgment interval of the expert judgment method reduces the errors of the random forest, different from the single expert judgment method, two times of expert judgment are applied, the five-element risk grade of the logistics system is obtained by the first time of expert judgment method, the second time of KNN algorithm is combined according to the records of the overall risks of the logistics system judged by the expert judgment method, the results obtained by the expert judgment qualitative analysis are quantitatively processed by the legal KNN, and the case analysis shows that the overall risk judgment model of the logistics system after the KNN training is different from the model set by the expert judgment method.

Drawings

Fig. 1 is a schematic flow chart of the epidemic prevention optimization method for epidemic infectious disease in a logistics system based on big data.

FIG. 2 is a schematic diagram showing the matching of the elements of the logistics system of the present invention.

FIG. 3 is a diagram illustrating the identification process of quarantine identification in the present invention.

FIG. 4 is a diagram illustrating the relationship between individual algorithms in the inventive combinatorial algorithm of the present invention.

FIG. 5 is a schematic diagram of the prediction principle of the random forest algorithm in the present invention.

FIG. 6 is a schematic diagram of the quarantine certification logic structure according to the present invention.

FIG. 7 is a conceptual structural diagram of the proof of quarantine according to the present invention.

FIG. 8 is a front view of the proof of quarantine of the present invention.

FIG. 9 is a chart of random forest regression tuning according to an embodiment of the present invention.

FIG. 10 is a graph of the degree of fit of the random forest regression test set in accordance with the present invention.

FIG. 11 is a reference map of the KNN model according to the embodiment of the present invention.

FIG. 12 is a visualization scatter plot according to an embodiment of the invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be obtained by a person skilled in the art without making any creative effort based on the embodiments in the present invention, belong to the protection scope of the present invention.

Referring to fig. 1, the application discloses a logistics system epidemic prevention optimization method based on big data, which comprises the following implementation steps:

1. analysis of elements of logistics system

In the environment of epidemic diseases, the transmission of viruses is caused by direct contact or indirect contact and the flow of virus mixtures along with air, so that the possibility of mutual infection exists among all elements in a logistics system. In consideration of the transmission characteristics of epidemic diseases, all elements in the logistics system are considered in the data collection range of the epidemic diseases, and mainly comprise five elements of drivers, goods packages, turnover tools and vehicles.

2. Epidemiological disease related key data collection

And according to the fact that whether the logistics system has a quarantine result or not, the logistics system is divided into a first entry logistics system and an entered logistics system, an entry quarantine port collects all information of the first entry logistics system, and after entry, each quarantine node obtains the passing geographical position information of the entered logistics system after entry through quarantine certification RFID information.

2.1 first entry Logistics System epidemic infectious disease Key data Collection

(1) Driver epidemic disease risk data acquisition and data source

The information of the driver is divided into three categories of basic information, health condition and epidemic disease risk assessment data according to the application field of the data. The basic information comprises name, gender, age and nationality, the health condition comprises body temperature condition, epidemic disease detection report and electronic health certificate, and the epidemic disease infection risk data comprises residence place, driving starting place, approach place, shipping area and delivery area and two types of container loading modes (logistics company loading and factory loading).

The basic information of the driver is derived from an identity card or a driving license; the body temperature condition in the health condition is derived from a body temperature measurer; the epidemic infectious disease detection report and the electronic health certificate are originated from a government epidemic infectious disease traceability platform, and if the government epidemic infectious disease traceability platform has no epidemic risk information, epidemic infectious disease detection information is provided through a paper or electronic version epidemic infectious disease pathogen detection report; the residence place, the driving starting place and the driving route are derived from the big data travel card, if no information of the big data travel card exists, the residence place is provided through the road operation license, the data of the starting area and the transportation area are provided by the import customs declaration form, and the container loading mode needs to be provided by the warehousing form. In order to obtain the data, the driver needs to provide documents such as an identity card, a driving license, a epidemic disease detection report, an electronic health certificate, a big data travel card, a road transportation operation license, an import customs declaration form, a warehouse entry form and the like, or wireless radio frequency codes such as two-dimensional codes, bar codes and the like which are authenticated by a national standard organization or other codes.

(2) Vehicle epidemic risk data acquisition and data source

And (1) dividing the vehicle data into basic information and epidemic infectious disease infection risk assessment data, wherein the basic information comprises vehicle type and entrance permit information, and the later data comprises a registration place, a license plate, a country of origin (region), an exit port, a loading port, an in-home goods source place and an approach place.

Wherein, the vehicle type, the entry permit information, the license plate, the exit port, the origin country (region), the loading port and the domestic cargo source are from the import customs declaration; the vehicle registration place is provided by a road transportation operation license; the approach is based on the driver's big data trip card. In order to obtain the data, the driver needs to provide documents such as import customs declaration form, road operation license, large data travel card and the like.

(3) Cargo epidemic disease risk data acquisition and data source

And (1) dividing the goods data into goods names, number information and epidemic infection risk assessment data, wherein the number information comprises commodity names, specification models, quantities and units, and the epidemic infection risk assessment data comprises goods production places and loading ports. The data all come from import customs clearance, so the document that the driver needs to provide is the import customs clearance.

(4) Epidemic disease risk data acquisition and data source of turnover tool

And (1) dividing the turnover tool data into basic information and epidemic infectious disease infection risk assessment data. The basic information comprises the model number of the turnover tool and the box number of the container, and the epidemic disease risk data of the turnover tool comprises the loading port and the route of the container. The turnover tool model and the container number, namely the loading port data, are derived from an import customs declaration and refer to a driver big data travel card in a way.

(5) Goods packaging epidemic disease risk data acquisition method

Since the goods package is generally performed in the production area and the logistics transfer center in the logistics activity, and the goods package is in contact with the environment outside the logistics system in the loading activity, the influence factors influencing the goods package are classified into 3 factors of the package type, the origin and the loading port. However, in the case where the container loading manner is the logistics company loading, the driver comes into direct contact with the package of the goods, and the type of the container loading is also taken into the judgment factor. Wherein the package type, origin and port data are derived from the import customs declaration. The driver needs to provide an import customs declaration.

2.2 Collection of important data of epidemic infectious diseases in the inbound Logistics System

When the imported logistics system passes through the domestic quarantine node, the RFID in the quarantine certificate sends out radio frequency signals, the RFID information reader-writer of the quarantine node receives the quarantine disinfection information and the path geographical position information of the logistics system, and the path geographical position information is used as a main basis for re-evaluating the epidemic infectious disease risk level of the logistics system.

The quarantine result authentication standard is as follows:

if the logistics system has the quarantine certificate, the quarantine result mutual recognition in the quarantine certificate is a standard established through visual data, and mainly comprises the following steps: (1) after the logistics system is authenticated in the immigration, the inspection of quarantine authentication is required to be carried out at the junction of two regions of the internal road, and if the logistics system does not pass through a high risk region, the elimination and killing are avoided. (2) According to the data retrieval of big data visualization, the high risk area in the administrative district is not existed in the whole transportation process, so that the elimination can be avoided when the user leaves the area. (3) If a new and unknown epidemic disease occurs, when entering the next logistics node, the logistics system risk information needs to be updated according to the new epidemic disease risk place, and new quarantine authentication is performed. (4) All methods, data, equipment and data transmission are completed under the safety condition of government certification standards.

2.3 epidemic disease Risk assessment database

The collected important data related to the epidemic infectious disease is stored in a temporary sample database in the epidemic infectious disease risk evaluation database, so that the subsequent matching work with other related databases in the epidemic infectious disease risk evaluation database is facilitated, and the introduction of the epidemic infectious disease risk evaluation database is as follows:

the epidemic disease risk assessment database D contains D₁、D₂、…、D_jAnd the sub-databases comprise a global geographical position epidemic disease risk grade database, a quarantine and disinfection database, an epidemic infectious disease pathogen environment detection result database, a human health database, a vehicle model database and the like. Wherein the data type code of the epidemic risk level of the geographic location is 1, namely D₁。

The risk index setting standard is based on the policy of dividing the epidemic infectious disease risk area in a country, namely, more than 50 accumulated cases occur, and the high risk area is the area with the occurrence of the aggregated epidemic infectious disease within 14 days. According to the population density of the country, the risk index is obtained by utilizing the epidemic infectious disease risk area division policy of the country, namely the risk grade index value of a certain foreign geographical position is as follows:

in the formula (2), NC is the number of newly-increased epidemic disease infectors in 14 days in the area, pd_LIndicating the population density of the area,

indicating the population density of the country.

3. Evaluation method

The invention adopts a random forest regression and expert judgment method and a KNN combined prediction algorithm, and takes the combined algorithm as the evaluation method of the invention. The algorithm principle is shown in fig. 4.

Establishing a epidemic infectious disease risk evaluation database by taking epidemic infectious disease risk condition information of various regions published by government departments as a data source, perfecting the epidemic infectious disease risk evaluation database by random forest regression, and matching the related information of the epidemic infectious disease of the logistics system with the epidemic infectious disease risk evaluation database by using a matching method (the epidemic infectious disease risk evaluation of the first-time entry logistics system uses the related data of the epidemic infectious disease acquired by an entry quarantine port, and the epidemic infectious disease risk evaluation of the entry logistics system uses the geographical position information of the route acquired by an entry logistics node after the entry), so as to obtain the epidemic infectious disease risk value of each element of the logistics system, and the risk grade of each element of the logistics system is taken as an input item to input a trained KNN model, and finally obtain the risk grade of the epidemic infectious disease of the whole logistics system.

(1) Random forest regression

A plurality of countries (hereinafter referred to as accurate countries) with the number of epidemic infectious disease infectors accurate to a first-level or lower-level administrative area are used as training samples, data of the number of newly added infectors in 14 days of each lowest-level administrative area of the countries, the proportion of population gathering activities, the population proportion, the GDP proportion, the proportion of passenger turnover to the country, the proportion of transportation cargo turnover to the country and human development indexes are combined into a data set, a certain proportion of data is extracted from the data set to serve as a training set, the rest of data serve as a test set, and a random forest regression method is used for obtaining a prediction model. The method comprises the steps of forming a data set by the population gathering activity proportion, the population proportion, the per capita GDP, the national proportion of the passenger turnover, the national proportion of the transport cargo turnover and the human development index of the lowest administrative level, which can be obtained by a non-precise country, substituting the data set as an input item into a prediction model to obtain an output item of the newly increased number of infectors in the national proportion of infectors, and thus obtaining the distribution situation of the infectors in all countries and provinces where the number of epidemic infectors cannot be precisely provincial. As shown in fig. 5.

The method for calculating each index value in the training sample set comprises the following steps:

TABLE 1 calculation method for each index value in training sample set

Basic model of random forest algorithm:

in the formula (3), F^*(x) For predicting the result, the model means an output data set F (x, θ) based on an independent variable x_i) Average of a decision trees.

The random forest algorithm process is as follows:

the first step is to select the proportion of newly-increased infectors in 14 days to national infectors as a characteristic variable Y and the rest as non-characteristic variables Y, and to use a leave-out method to collect the accurate national original data set { M_m+nRandomly extracting 80% of data as a training sample set

The remaining 20% were used as test set

And secondly, randomly extracting tau features from the six features of the training sample set, wherein the extracted tau features are not changed in the random forest generation process, and the features which are not extracted are subjected to node splitting.

And thirdly, circulating the first step and the second step, and adjusting parameters to finally obtain the optimal prediction model belonging to the random forest.

The fourth step, test set data

The ratio of population gathering activity, population proportion, GDP ratio, passenger turnover to national ratio, transport cargo turnover to national ratio and human development index of each administrative region are used as input items, the ratio of newly-increased number of infectors in each administrative region at 14 days to national infectors is used as an output item, and a test prediction set is obtained

Test prediction set and test set

And comparing the real samples. The actual value data set of the proportion of newly-increased infectors to nationwide infectors within 14 days of each province in the test set is

The prediction utility discriminant is:

in formula (4), S_EVSRepresenting an interpretive variance regression score, var {. Denotes variance.

When the explained variance regression score is not less than 0.9, the prediction model is considered to be valid.

Fifthly, forming the non-characteristic variable Y data of each region of the non-precise country into a data set

Substituting the prediction model epsilon to obtain a prediction data set containing k data

Predicting a data set

The number of newly-increased infectors in each predicted value and 14 days of non-precise country

Multiplying to obtain a new number data set of 14 newly-increased infectors in the precise country within 14 days of each administrative district

Will be provided with

And

transverse consolidation into data sets

(non-precision country), will { M_m+nThe number of newly-increased infectors in 14 days after the characteristic variable y is changed is obtained as a data set

(accurate country) with

The data set { M is obtained by transverse combination_m+n+kAs a preferred scheme, a data set { M } is extracted_m+n+kObtaining a data set y by the characteristic variable y in the data set_m+n+k}，{y_m+n+kSubstituting the geographical position risk grade index into a judgment formula (2) to obtain a data set

Data set

And { M_m+n+kLongitudinally combining to obtain a geographical position risk value data set (M) of each state administrative district_m+n+kThe data set is used as a geographic position epidemic disease risk value database D₁。

[ Note ]: the rows are added by horizontal combination, the columns are added by vertical combination, and the subscript in the data set represents the number of samples, i.e., the row number.

(2) Matching method

And matching the collected key data related to the epidemic of the logistics system with the risk evaluation database of the epidemic of the logistics system so as to obtain the risk level of each element in the logistics system.

The matching model is as follows:

map(i_jk,D_j),(i＝α、β、γ、λ、Ω) (5)

in the formula (5), alpha, beta, gamma, lambda and omega respectively represent a driver, a vehicle, a turnover tool, packaging and goods, map (-) represents a matching function, i_jkK-th data in j data type representing element i, D represents epidemic disease risk database, D_jThe data type corresponding to the j data type of the element i in the epidemic disease risk database is shown, and a matching schematic diagram is shown in fig. 2.

According to the characteristic that the epidemic infectious diseases are easy to spread in a short distance, the transportation tool is divided into a cab and a goods area, the turnover tool is divided into an inner wall and an outer wall, and the package is divided into the inner wall and the outer wall. Five elements of the logistics system are analyzed according to contact types, and the elements for generating close contact are as follows: the outer walls of the package and the inner walls of the transfer tool, the cab and the driver of the vehicle, the driver and the outer walls of the package (when the container is packed in a logistics company).

In equation (6), max (. Cndot.) is a function of the highest risk value,

the risk created by the i element's path through each geographic location,

the epidemic position risk level expression of other logistics system elements which are in close contact with the i element,

representing the risk level represented by the environmental test result of the epidemic infectious disease pathogen of the i element.

(3) Logistics system risk determination method based on expert judgment method and KNN algorithm

The scoring rules are as follows:

the scoring rule comprises 4 indexes of the survival rate (1-lethality rate) of the epidemic infectious disease, the non-infection rate (1-infection ratio), the degree of influence on public safety and the degree of influence on social economy, the weights of the four indexes are determined by an expert group together, the infection rate and the lethality rate of the epidemic infectious disease are determined according to the actual situation of the infectious disease, the expert group scores the risk grade according to the degree of influence on the public safety and the degree of influence on the social economy of the epidemic infectious disease, and the scoring range is 0-1.

The expert scoring table is as follows:

TABLE 2 expert scoring table

In the expert rating Table, X₁To X₄The weight of 4 indexes of the infection rate, the fatality rate, the degree of influence on public safety and the degree of influence on social economy of the epidemic infectious disease, H_i1To H_i4High risk judgment values for 4 indexes, respectively, for i experts, where H_i1、H_i2For unchangeable fixed values of expert, H_i3、H_i4Scores given to experts, medium risk, low risk correspondence index scores and high riskSimilarly. Weighting yields i expert scores for three risk levels, as follows:

the scoring profile for the panel was:

TABLE 3 expert group Scoring

Wherein:

in the formula, L_iIndicating the recommended value of the score for the ith expert for a low risk level, M_iIndicating the proposed value of the score for the ith expert for a medium risk rating, H_iAnd the system comprises a plurality of experts, wherein the experts are represented by scoring recommendation values of the ith expert for high risk levels, n represents the number of people in the expert group, L represents a finally adopted low risk level value, M represents a finally adopted medium risk level value, and H represents a finally adopted high risk level value.

The risk assessment value delta of epidemic pathogen carried by the logistics system is obtained through the steps_RAccording to the index delta_RThe risk of epidemic pathogens carried by the logistics system is classified into the following grades: delta_R∈[L,M]Logistics systems are low risk; delta_R∈[M,H]The logistics system is in danger; delta_R∈[H,1]Logistics systems are high risk.

The risk levels of the respective epidemic infectious diseases of the five elements of the logistics system are obtained through a matching method, the scores of the low, medium and high risk levels obtained through an expert judgment method are used for assigning the five elements of the logistics system, and a risk matrix X belonging to the logistics system can be obtained^1,5And inputting the risk matrix into a KNN algorithm to obtain the overall risk grade of the logistics system.

KNN (K-nearest neighbor) algorithm principleIn order to analyze the characteristic value of the characteristic variable F to be researched at K points (K values) nearest to the sample Y in a plane, space or high-dimensional space, a tree is constructed according to the K values, the maximum number of the constructed trees is set, and a certain characteristic value F with the maximum occurrence frequency in the characteristic variable F at the K points is recorded^*And F is^*As the value of the sample Y characteristic variable F. The K value is the limited radius of the KNN, namely K points closest to the sample are selected as a judgment basis of the characteristic value of the sample; the distance calculation method comprises three methods, namely a Manhattan distance, an Euclidean distance and a Chebyshev distance, and the distance formula is as follows:

in the formula (8), L_p(x_i,y_j) Denotes x_iAnd y_jThe distance between the two points is such that,

l when p is 1_p(x_i,y_j) For Manhattan distance, L when p is 2_p(x_i,y_j) At Euclidean distance, L when the p region is infinite_p(x_i,y_j) Is the chebyshev distance; constructing a tree according to a distance formula, wherein the type of the constructed tree comprises brute force realization, KD tree realization and ball number realization, the brute force realization is suitable for the condition that a data set is distributed more dispersedly, the KD tree is suitable for a space with the dimensionality smaller than 20, and the ball number is suitable for a space with the dimensionality larger than 20; the maximum number of the construction trees is a threshold value of the number of leaf nodes for stopping building the tree, and is an upper limit of the construction tree construction.

The epidemic infectious disease risk values of five elements of the logistics system can be obtained by a matching method and an expert scoring method, and the five risk values can form a sample matrix [ R ]_α,R_β,R_γ,R_λ,R_Ω]And the following steps are carried out by taking the data as a data set to be predicted:

firstly, judging the integral risk level of n logistics systems according to an expert judgment method to be used as a historical record, and listing the risk level of five elements of the logistics systems and n historical records of integral risk judgment into a matrix X_t∈R^n,6Extracting the overall risk level (low risk, medium risk and high risk) of the logistics system from the training sample data set as a data label, and randomly extracting 80% of data from the training sample data set as the training sample data set by adopting a leave-out method

i belongs to {1,2, \8230;, m }, and the sample set formed by the data tags in the training sample set is { y ∈_iThe sample set composed of five elements of the logistics system is { Y }_i}, the remaining 20% as test set

The sample set composed of the data labels in the training sample set is

The sample set composed of five elements of the logistics system is

Second, will { y_iAnd { Y }_iAnd inputting the KNN model to train to obtain the KNN model omega.

Thirdly, test set data is acquired

Respective risk grade of five elements of medium logistics system

As an input item, the overall risk level of the logistics system is used as an output item to obtain a prediction set of the overall risk level of the logistics system

Will test the prediction set

And test set

Middle true sample

And inputting the ACC model to obtain a classification accuracy score. The prediction utility discriminant is:

ACC≥0.9 (9)

in the formula (9), ACC represents the classification accuracy score.

And when the classification accuracy score is not less than 0.9, the prediction model is considered to be effective.

And fourthly, adjusting parameters in the KNN model by the ACC to enable the ACC to reach the maximum value, namely, the model to reach the optimum value, and accordingly obtaining the prediction model omega. K, undersfit, overfit.

Fifthly, substituting the data to be predicted into the prediction model omega to obtain a prediction data set containing 1 piece of data

This data is taken as the overall risk level of the logistics system.

4. Quarantine method

And carrying out different sampling inspection treatment on logistics systems with different risk grades, quarantining all five elements in the logistics systems when the logistics systems are in a high risk grade, quarantining the medium risk and the elements above in the logistics systems when the logistics systems are in a medium risk grade, quarantining the medium risk and the elements above when the logistics systems are in a low risk grade, and randomly extracting any one of other elements for quarantine.

When the risk level of the driver is low, the driver can be free from inspection, and when the risk level of the driver is higher than the middle risk level, the driver can be subjected to epidemic infectious disease inspection. And when the risk grades of the vehicles and the turnover tools are at or above the medium risk, the interiors and exteriors of the drivers' cabs of the vehicles and the turnover tools are quarantined. When the risk grade of the goods and the package is low risk, 2 percent of random sampling inspection is adopted, when the risk grade is medium risk, random spot inspection is carried out at a ratio of 20%, and when the risk level is high, all goods and packages need to be quarantined.

And when the epidemic pathogens are not found in the sample detection, the judgment of the risk level is considered to be correct, and a normal disinfection process is carried out. When the sample is detected to find out the epidemic infectious disease pathogens, the logistics system is immediately changed into a high risk grade, the logistics system is transferred and isolated, the driver is isolated and observed, and other elements are comprehensively sterilized.

5. Method for killing bacteria

In the technical Specification for disinfection and sterilization in China, the disinfection process is divided into four disinfection grades, namely a disinfection method for killing all microorganisms, namely a disinfection method by a disinfection method, a high-level disinfection method, a medium-level disinfection method and a low-level disinfection method according to the level of the action of a disinfection factor; the high-level disinfection method is a disinfection method for killing all bacterial propagules, viruses, fungi and spores thereof and most bacterial spores; the middle level disinfection method refers to a disinfection method for killing microorganisms outside bacterial spores; low level disinfection refers to a disinfection method that kills bacterial propagules and lipophilic viruses. Because epidemic diseases are caused by viruses, low-level disinfection methods are abandoned, and sterilization, high-level disinfection methods and medium-level disinfection methods are adopted.

The sterilization method adopts 5 times of diluent of the compound hydrogen peroxide disinfectant to spray on the surface of an object, and the effect is carried out for 5 minutes; the high-level disinfection method adopts 0.1 percent sodium hypochlorite solution to be sprayed on the surface of an object and acts for 5 minutes; the medium level disinfection method adopts 75% ethanol to spray on the surface of the object, and the action lasts 10 minutes.

TABLE 4 disinfectant level method table

Level of disinfection	The disinfectant is prepared from	Action time (minutes)
			Method of sterilization	5-time diluent of compound hydrogen peroxide disinfectant	5
High level disinfection method	Sodium hypochlorite solution 0.1%	5
			Middle level disinfection method	75% ethanol	10

And (4) according to the confirmed risk grade, matching a known disease source with an unknown disease source to kill the disease source, and killing the disease source. In the case of unknown epidemic pathogen or spore-free pathogen, the elements of the logistics system with medium-high risk level except drivers adopt a sterilization method, and in the case of spore-free epidemic pathogen, the following sterilization method is adopted:

in order to reduce the risk grade evaluation deviation of each element caused by unexpected cross infection in the logistics system, different disinfection methods are adopted for 5 elements according to the risk grade of the logistics system, when the system is at high risk, the high risk grade element in the system adopts a disinfection method, and the medium and low risk grade elements adopt a high-level disinfection method; when the system is in the middle risk, the high-level disinfection method is adopted for the high-level and middle risk grade elements in the system, and the low-level disinfection method is adopted for the low-risk grade elements; when the system is at low risk, the high risk grade elements in the system adopt a high level disinfection method, and the medium and low risk grade elements adopt a medium level disinfection method.

As can be seen from the 5-element risk assessment rule, when the goods are at a high risk level, the package is determined to be at a high risk level, but the package may be infected by the driver because the container loading manner is that of the logistics company, so that the risk level of the package is equal to or higher than that of the goods. Due to the characteristics of packaging and wrapping the goods, the package and the goods are disinfected as a whole. When the goods risk level is low risk, only the package is sterilized; when the goods risk level is medium-high risk, the goods are withheld and virus detection is carried out, and the package and the goods are disinfected. The quantity of the goods can be known through the goods epidemic disease risk database so as to select a proper disinfection mode and obtain the expected disinfection time.

Because the turnover tool and the vehicle are relatively difficult to disassemble, and the cab of the vehicle is in direct contact with a driver, the vehicle is divided into two parts, namely the cab and the appearance. The turnover vehicle and the vehicle are taken as a whole, different disinfectants are respectively sprayed according to respective risk grades when the risk grades are different. The driver is disinfected individually and the cab is disinfected individually.

6. Evaluation of killing Effect

Before the logistics system is killed, an aqueous solution containing GFP marked adenovirus with a certain concentration is sprayed on the surface of the logistics system, and the adenovirus is consistent with the environment of each element in the system, so that the killing rate of virus carried by the system is equal to the killing rate of the adenovirus. And after the adenovirus solution is sprayed, collecting the sample on the surface of the object, calculating the amount of adenovirus, and after killing, collecting the sample on the surface of the object again, and calculating the amount of adenovirus.

In the technical Specification for killing bacteria in China, the standard of the killing rate of microorganisms in air is 99.91 percent, so the killing rate is adopted as the qualified index of killing, and when the killing rate is more than 99.91 percent, the killing is qualified. The kill rate formula is:

in the formula (10), C is the virus killing rate, T₀The number of labelled adenoviruses detected in the pre-kill sample, T₁Representing the number of labeled adenoviruses detected in the post-kill sample.

7. Proof of quarantine

In order to prove that the processed logistics system has safety, the accurate geographical position of the logistics system path after entry is recorded, so that the domestic quarantine node can carry out accurate epidemic risk assessment again after the logistics system enters, the quarantine time and cost increased by the quarantine and disinfection again after the logistics system enters are reduced, the logistics system certified material after the quarantine and disinfection is given at the entry port, and the data is stored for retrieval. In the aspect of the design of the quarantine result certification mark, the invention mainly relates to a quarantine certification terminal, an RFID information base and anti-counterfeiting verification.

(1) Quarantine certification terminal

The appearance of the quarantine certification terminal is shown in fig. 7, in which: the quarantine certification terminal front face 1 specifically shows fig. 8, and the information includes certificate number and certificate control unit information, and charge port 2 is the battery charge end, and electro-magnet 3 is used for adsorbing on the commodity circulation vehicle surface. The quarantine certification terminal comprises a storage battery, a charging port, an electromagnet, a navigation positioning chip, a memory and an active RFID, the connection principle of all the components is shown in figure 6, the storage battery supplies power for the electromagnet, the navigation positioning chip, the memory and the active RFID, the positioning navigation chip stores position information into the memory, the quarantine node exchanges epidemic disease risk information of the logistics system with an RFID reader-writer through active RFID transmission after entering the border, and finally the information is transmitted to an epidemic disease risk evaluation database for evaluation of the quarantine node.

The battery adopts the lithium cell, and the port that charges adopts USB Type-C interface, and the electro-magnet adopts Shenzhen market nine article magnetism science and technology Limited's JNP-15/15, and navigation positioning chip adopts big dipper location module, and the memory adopts the tf card, and active RFID adopts hyperfrequency RFID.

(2) RFID information base

The RFID information base comprises three types of information, a Part1 comprises a label making unit, making time and a record number, a Part2 comprises risk information, risk grades, a killing mode and a killing effect of each element of the logistics system, and a Part3 comprises geographical position information of the logistics system after entering the border. The information base is as follows:

TABLE 5 RFID storage information Table

Referring to fig. 3, in the information base, the killing effect, the killing mode and the killing effect of Part1 and Part2 are directly displayed after the RFID tag is read, part2 risk information is not displayed and is imported into the quarantine killing database, and Part3 geographical location information is not displayed and is imported into the quarantine killing database after entering the country. And judging whether the logistics system is quarantined and disinfected or not by judging whether the logistics system has RFID certification or not by the logistics system at each level of logistics nodes entering the logistics port and entering the logistics port, and if not, quarantining and disinfecting the logistics system and giving the logistics system RFID certification after the disinfection is finished. If the logistics system has RFID certification, the logistics system risk is re-judged according to the updated risk information (namely the passing area risk level) of the logistics system, which is not contained in the RFID certification information, and the vehicle is released when the updated logistics system risk information is low risk; if the risk is not low, quarantine and disinfection are carried out again, and information contained in the RFID is updated.

(3) Quarantine certificate anti-counterfeiting verification

The docketing number in the RFID information base Part1 is a certificate for verifying whether the RFID is valid, and the logistics quarantine node judges whether the RFID is valid by searching whether the docketing number of the logistics system exists in a quarantine disinfection database. The record number is encrypted through a secret key, is displayed as the encrypted record number when the logistics node reads the RFID, and is decrypted and compared with the recorded record when the logistics node is imported into the quarantine disinfection database.

The following will be further illustrated by taking an epidemic of an infectious disease as an example:

example (b):

background: 24/10/2020, the regional epidemic situation of a certain infectious disease in a certain region of country A is outbreaked, after tracing the source of the virus, the disease control center discovers highly homologous virus with the virus in the region when extracting border-passing containers of three adjacent countries in the region, and determines that the epidemic situation of the infectious disease in the region comes from the overseas containers of the adjacent countries.

The method comprises the following steps: scene reproduction method

Now, selecting a port a closer to a certain region of the country A as an implementation port, selecting a general commercial country B of the port a as a logistics system research object, and processing a logistics system from the country B on the port a by using the method as follows:

1. epidemic risk database processing

Within 14 days 24 before 10 and 2020, the data for a newly-increased infector of a certain epidemic situation in country B is as follows:

TABLE 6 data of infected persons of certain infectious disease epidemic situation of country B

And taking the data of the first-class administrative regions of the seven countries of the country M, the country N, the country P, the country Q, the country R, the country A and the country T as the model training data set data.

Sampling from the data by using a leave-out method, wherein a sampling proportion training set and a test set are 8:2, the sampling freedom degree is 15, and a training set is obtained

And test set

Inputting the training set into a random forest model, circularly traversing possible values of parameters of the random forest model to obtain an optimal parameter combination when the regression score of the interpretation variance S of the random forest model reaches the maximum value, wherein the optimal values of the parameters are shown in a table 7:

TABLE 7 random forest model parameters

Inputting the training set into a training model, and after the model training is finished, testing the set

The relevant indexes in the model are input to obtain the prediction, and a prediction set is tested

Testing prediction set and testing set characteristic variable real data set

A comparison is made. Finally obtain S_EVS＝0.9595175953737675>0.9, the random forest model is effective, and the fitting degree of the random forest test set is shown in figure 10, so that the fitting degree is better.

The data set to be predicted consisting of country B-related data is as follows:

TABLE 8. Country B dataset to predict

Prediction data set composed of prediction results

As shown in table 9, y _ test _ pred represents the proportion of newly-added infectors to nationwide infectors within 14 days of national B-stage administrative district predicted by random forest model. The newly increased number of infected persons in the first-class administrative district of the country B within 14 days can be obtained by the data of the infected persons of the epidemic situation of the certain infectious disease in the first 14 days of the country B in the table 6 and the forecast data set of the country B in the table 15, and the newly increased number of infected persons in the first-class administrative district of the country B within 14 days can be obtainedThe number of dyers is calculated by selecting the smallest integer greater than the number of dyers. According to the formula (2), the epidemic infectious disease risk value of each stage of administrative district of the country B can be obtained:

TABLE 9 national B prediction data set

2. Data acquisition and matching for logistics systems

According to the main commodity type of the imported country B of the country A and the domestic industrial structure and market structure of the country B, possible logistics system information from the country B is matched with a global geographic position epidemic risk grade database, an epidemic infectious disease pathogen environment detection result database and a human health database in an epidemic infectious disease risk evaluation database to obtain risk values of each data of five elements of the logistics system, wherein the risk values are shown as follows and are shown in tables 10-14:

TABLE 10 information gathering table of the persons who sample the cargo along with the immigration

TABLE 11 sample entry vehicle information collection sheet

TABLE 12 sample inbound goods Collection information sheet

Table 13 sample immigration turnover tool information collection table

TABLE 14. Sample inbound goods packing collection information sheet

2. Logistics system data matching

According to tables 10-14, the risk values for five elements of the logistics system can be obtained:

R_α＝max(0,0,0,0.99,0.99,0.99,0.99,0.99,0)＝0.99，

R_β＝max(0.99,0.99,0.99,0.99,0.99,0.99)＝0.99，

R_γ＝max(0.99,0.99,0.99,0.99)＝0.99，

R_λ＝max(0.99,0.99,0.99)＝0.99，

R_Ω＝max(0.99,0.99,0.99)＝0.99。

according to the contact condition of five elements of the logistics system, the following formula (6) can be obtained:

R_β＝max(R_β,R_α)＝0.99，

R_γ＝max(R_γ,R_Ω)＝0.99。

prediction set data composed of respective risk values of five elements of logistics system

Is [0.99 0.99 0.99.99.0.99.99]。

3. Five-element and overall risk level determination of logistics system

Risk assessment value delta of epidemic pathogen carried by logistics system obtained by expert judgment method_RComprises the following steps: delta_R∈[0,0.3]Logistics systems are low risk; delta_R∈[0.3,0.9]The logistics system is in danger; delta_R∈[0.9,1]Logistics systems are high risk.

Setting 5 columns of random numbers with 1000 rows, wherein 5 columns respectively represent risk values of personnel, vehicles, turnover tools, goods packages and goods, and the value interval of 80 percent is [0,0.2 ]]The value interval of 16% is [0.2,0.8 ]]The 4% value interval is [0.8,1']The 5000 values are scrambled into a matrix X_t∈R^1000,5Where each row represents one sample data. For matrix X_t∈ R^1000,5The following treatment is carried out: (1) randomly selecting 200 data in the matrix, setting a turnover tool loading mode as a logistics company loading mode, considering that the personnel and the goods package are in contact according to the contact condition of five elements of a logistics system, and giving the highest risk value of the driver and the goods package to the personnel and the goods package. (2) And according to the contact condition of the five elements of the logistics system, taking the highest risk value of the personnel and the vehicle, and giving the highest risk value to the vehicle, wherein the risk value of the personnel is unchanged. (3) According to the contact condition of five elements of the logistics system, the highest risk value of the turnover tool and the goods package is given to the turnover tool, and the risk value of the goods package is unchanged.

Assuming an expert judgment rule, obtaining the risk grades of 5000 numerical values in the historical data set according to an expert judgment method, wherein the overall epidemic infectious disease risk grade rule of each sample logistics system is set as follows: (1) and when the five-element risk levels of the logistics system are all low risk, the risk level of the epidemic disease of the whole logistics system is considered to be low risk. (2) When one risk level of the five elements of the logistics system is the intermediate risk with the risk levels of the two elements, the risk level of the epidemic infectious disease of the whole logistics system is considered as the intermediate risk. (3) When three element risk levels of five elements of the logistics system are medium risks, the risk level of the epidemic infectious disease of the whole logistics system is considered to be high risk. (4) And considering the risk level of the epidemic disease of the whole logistics system as high risk when one or more than one element risk levels in the five elements of the logistics system are high risk. According to the rule, the risk grades of all sample logistics systems can be obtained, the low risk is set as a label 2, and the stroke is carried outRisk is set to label 1 and high risk is set to label 0, resulting in matrix X_t∈ R^1000,6And taking the data as a historical data set of the KNN algorithm.

80% of historical data set is taken as training sample data set by using a leaving method

The rest is used as a test set

The overall risk level of the logistics system is extracted from the training sample data set to serve as a data label, the KNN model is used for training the training sample data set, the optimal parameter combination when the ACC reaches the maximum value is obtained by a circular traversal setting-out method and the possible values of parameters of the KNN model, and the optimal values of the parameters are shown in the table 15:

TABLE 15 optimal values of parameters of KNN model

The final classification accuracy score obtained is: ACC =0.975, and it is found from formula (9) ACC >0.9 that KNN model ω is effective.

Data to be predicted

Substituting the prediction model omega to obtain a prediction data set

The prediction result is [0 ]]That is, the epidemic risk level of the study subject of the logistics system from country B is a high risk level, the scatter diagram is shown in fig. 12, the dots in the scatter diagram represent the historical data sets, the stars represent the study subject of the logistics system, the white represents a low risk level, the gray represents a medium risk level, and the black represents a high risk level. As the score of the test set is 0.975, the KNN quantitatively processes the result obtained by the qualitative analysis of expert judgment, and the definition value of the expert judgment method is blurred, so that the model is more flexible.

4. Method for determining five-element quarantine method of logistics system

The rules determined by the quarantine method can be obtained, and the five-element quarantine method of the logistics system research object is shown in table 16:

TABLE 16 quarantine method of each element of logistics system research object

5. Method for determining five-element killing of logistics system

The rules are determined by a killing method, and the five-element killing method of the logistics system research object is shown in table 17:

TABLE 17 method for killing each element of research object of logistics system

And spraying a water solution of GFP marked adenovirus with a certain concentration on four elements of vehicles, turnover tools, goods packages and goods before sterilization, wiping the surfaces of all parts by using quarantine swabs, and recording the quantity of the GFP marked virus. After killing, wiping the surfaces of all parts by using a quarantine swab again, recording the quantity of the viruses marked by the adeno GFP, and judging the killing effect according to the formula (10). If the sterilizing effect does not reach the standard, sterilizing.

6. Proof of quarantine

And if the sterilizing effect of each part of the logistics system reaches the standard, giving an RFID quarantine result certificate to the logistics system. The RFID quarantine result proves that the information is as follows:

TABLE 18 RFID quarantine results certification information sheet

The foregoing shows and describes the general principles, principal features, and advantages of the invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments, and the above embodiments and descriptions are only preferred examples of the present invention and are not intended to limit the present invention, and that various changes and modifications may be made without departing from the spirit and scope of the present invention, which fall within the scope of the claimed invention. The scope of the invention is defined by the appended claims and equivalents thereof.

Claims

1. A logistics system epidemic prevention optimization method based on big data is characterized by comprising the following steps:

data acquisition: collecting epidemic disease risk data of elements of a logistics system;

matching killing grades according to the overall risk grade, and killing;

evaluating the killing effect;

2. The epidemic prevention optimization method for epidemic infectious disease in logistics system based on big data as claimed in claim 1, wherein the elements of logistics system include at least two of the following: driver, goods packing, turnover tool and vehicle.

3. The epidemic prevention optimization method for epidemic infectious disease in logistics system based on big data as claimed in claim 1, wherein said epidemic risk assessment database is derived from epidemic risk situation information published by government departments in various regions.

4. The logistics system epidemic prevention optimization method based on big data as claimed in claim 1, wherein before data matching, the number of epidemic infectors in precise countries is used as training samples, a random forest regression algorithm is adopted to obtain a prediction model, and the prediction model is adopted to process a risk database of non-precise countries epidemic infectious diseases to obtain a risk value database of geographical position epidemic infectious diseases.

5. The epidemic prevention optimization method for epidemic infectious diseases in logistics system based on big data as claimed in claim 1, wherein in data evaluation, the scores of low risk, medium risk and high risk grade are obtained by expert judgment method, and the elements of logistics system are assigned to obtain the risk matrix of logistics system, and the risk matrix is input into KNN algorithm to obtain the overall risk grade of logistics system.

6. The epidemic prevention optimization method for epidemic disease of logistics system based on big data as claimed in claim 1, wherein based on confirmed risk grade, matching killing method for known and unknown disease sources is carried out, and killing is carried out.

7. The epidemic prevention optimization method for epidemic infectious disease in logistics system based on big data as claimed in claim 1, wherein the evaluation of disinfection effect adopts: before killing, spraying the aqueous solution containing GFP-labeled adenovirus onto the surface of elements of the logistics system, collecting surface samples, calculating the amount of adenovirus, after killing, collecting the surface samples of the objects again, and calculating the amount of adenovirus.

8. The epidemic prevention optimization method for epidemic infectious diseases in logistics system based on big data as claimed in claim 1, wherein said quarantine certification terminal comprises wireless radio frequency information base, said wireless radio frequency information base comprises tag production unit, production time, record number, and risk information, risk grade, disinfection mode, disinfection effect of each logistics system element.

9. The epidemic prevention optimization method of epidemic infectious disease of logistics system based on big data as claimed in claim 8, wherein the docket number is encrypted by a key.