A medicine storage and taking system and device for a household intelligent medicine box based on a greedy strategy

Abstract

The application discloses a kind of based on the home intelligent medicine box medicine storage and medicine taking system of greedy strategy, it is characterized by comprising the following contents:(1) medicine storage: obtaining medicine information, then according to the prescription correlation degree calculation obtained storage box;(2) medicine taking: user needs to enter the identity information of medicine taker;User enters prescription, and the system automatically associates the medicine storage distribution in medicine box according to the above information;After according to the effective prescription entered, the system automatically arranges the time of user medicine taking, when to take medicine time, the system prompts through voice;After user authentication, the system automatically matches the name and quantity of medicine in prescription according to the identity identified, and judges whether there is missing medicine, if there is, then prompt;For when the system detects that user takes medicine according to prescription, the system pops up corresponding medicine grid.

Description

Technical Field

[0001] This invention relates to a home-use smart medicine cabinet system and device for storing and retrieving medicines based on a greedy strategy, belonging to the field of smart home products. Background Technology

[0002] Most home medicine cabinets on the market are physical cabinets, only designed for simple storage of medications. They offer no guidance on the information contained in the cabinet or on patient medication use. While existing smart medicine cabinets offer some intelligent functions, they still have significant shortcomings in terms of medication storage and categorization. They simply store medications and remind users to retrieve and take them according to regulations. However, they often fail to store medications systematically, with different types of medications in different compartments, making it difficult to match prescriptions with actual medications.

[0003] While the device of this invention is similar in appearance to other smart medicine boxes, it differs greatly in its purpose and technical implementation. The purpose of this invention is to make the storage of medicines in home smart medicine boxes more reasonable, standardized, and user-friendly. By combining algorithms such as greedy strategies and local search, the storage strategy for medicines is optimized. It can make efficient use of medicine box space according to user habits, especially providing the optimal medicine box storage in situations where prescriptions and usage habits are complex. Summary of the Invention

[0004] The purpose of this invention is to provide a home-use smart medicine cabinet system and device based on a greedy strategy for storing and retrieving medicines. This addresses the significant shortcomings in existing technologies regarding medicine storage and classification. These technologies simply store medicines and remind users to retrieve and take them according to regulations. However, they suffer from problems such as inconsistent storage, different types of medicines in each medicine box, and difficulty in matching prescriptions with medications.

[0005] The technical solution of the present invention is: a home smart medicine cabinet medicine storage and retrieval system based on a greedy strategy, comprising the following: (1) Medicine storage: obtaining medicine information and then calculating the storage box according to the prescription correlation; (2) Retrieval: the user needs to enter the identity information of the medicine user, including name and facial feature information, the system performs face recognition and records; then the user enters the prescription on the mobile APP, including the prescription owner, medicine name, dosage and time of administration; the system automatically associates the medicine storage distribution in the medicine cabinet according to this information; then the system automatically arranges the time for the user to retrieve the medicine according to the entered valid prescription, and the system will prompt the user with sound when it is time to take the medicine; then, the user can verify his identity through face recognition by camera, and the system will automatically match the name and quantity of medicine in the prescription according to the identified identity, and determine whether there are any missing medicines, and if so, prompt the user; when the system detects that the user retrieves the medicine according to the prescription, it will select the box where the medicine is located according to the following strategy: since a medicine may be placed in more than one box, in order to make it more convenient for the user to retrieve the medicine, the system will pop up as few boxes as possible while satisfying the prescription.

[0006] The drug information includes: drug ID, drug name, production date, shelf life, instructions for use, box length, box width, and box height.

[0007] If newDrug is included in the historical drug usage data, the number of days of use L is recorded. days Number of uses L freq Total amount of medicine used (L) num Given the recommended number of times drug i is taken per day, R f And the recommended dosage R n Define the use of the drug frequency function:

[0008]

[0009] Here, w1, w2, and w3 are weighting coefficients, and w1 + w2 + w3 = 1. We set w1 = 0.5 and w2 = 0.3.

[0010] w3 = 0.2; For the i-th medicine box, iterate through all the medicines in the medicine box and calculate the total usage frequency F. j :

[0011]

[0012] Step (1) also includes: considering centralized storage of prescriptions: selecting target boxes: selecting boxes that meet the volume requirements and have the highest prescription correlation, j* = argmax_j(A[j]), if there are multiple selections, then considering adjacency and utilization of remaining space; if there are multiple boxes that meet the conditions, further selection is based on the number of medicines with the same prescription in adjacent boxes, and a function can be defined. Choose the box with the largest value in C[j].

[0013] Step (2), define the decision variable x i (i = 1, 2, ..., m) are binary variables, where m is the number of cells containing the medicines in the prescription; if box i is selected, then x i =1, otherwise x i =0, so that the objective function minimizes the number of boxes used. Furthermore, for each drug j (j = 1, 2, ..., n), a constraint is added to ensure that it is selected at least once. Where S i Let be the set of boxes containing medicines in the i-th box. By iterating through each medicine and checking which boxes contain it, and ensuring that at least one of these boxes is selected, constraints are constructed to find as few boxes as possible that satisfy the prescription. The system then automatically dispenses the medicine and checks whether the medication collection is complete. If the user fails to collect the medicine on time, the system will issue a timeout reminder. When the user has collected the medicine, the system will update the medicine information and remaining quantity in a timely manner to ensure the timeliness and accuracy of the information.

[0014] Step (1) also includes: space utilization optimization: on the basis of satisfying the above conditions, for all possible boxes, select the box with the smallest remaining space but still able to accommodate new medicines, so as to reduce space waste; define R[j] = B[j] / Capacity[j] as the space utilization of the current box, and select the box with the smallest R[j].

[0015] Step (1) also includes: calculating the prescription correlation degree:

[0016]

[0017] Where: A[j] represents the prescription relevance score of box j; S is the set of all known prescriptions; N is the set of currently existing medicines in box j; i represents the i-th medicine, and k represents the k-th prescription; W k Indicates the degree of association between multiple drugs appearing in the same prescription; V i This represents the correlation between a drug appearing in multiple prescriptions; here, a function f(i,k) is defined such that if drug i exists in prescription k, then f(i,k) = 1, otherwise...

[0018] f(i,k) = 0;

[0019] (1) Formula for calculating the correlation between multiple drugs appearing in the same prescription:

[0020]

[0021] G was used here. k The quadratic function as the prescription S k The weighted correlation reflects the higher the correlation bonus when more drugs appear in the same prescription.

[0022] (2) Formula for calculating the correlation of a drug appearing in multiple prescriptions

[0023]

[0024] H is used here. i The logarithm of the formula is used as the weighted correlation degree of drug i; the more prescriptions i appears in, the higher the correlation degree. In this prescription correlation degree calculation, emphasis is placed on the situation where multiple drugs appear in the same prescription, and these cases are given higher scores.

[0025] Step (1) also includes: Formula for calculating the concentration of prescriptions in adjacent boxes: This aims to assess how well the medicines stored in a box and its adjacent boxes meet prescription needs. This helps maintain the physical proximity of related medicines, making it easier for users to quickly find and use all the medicines in the entire prescription; Formula for calculating the concentration of prescriptions in adjacent boxes:

[0026]

[0027] Where: C[j] represents box j and its adjacent boxes j-1 and j+1, assuming proper handling of boundary conditions, the concentration score of prescriptions for boxes that are connected end to end or whose boundary boxes only consider the prescriptions of one side of their neighbors; S is the set of all known prescriptions; M[j] is the set of medicines stored in box j, which also includes the prescription information to which these medicines belong; I(s,M[j]∪M[j-1]∪M[j+1]) is an indicator function that evaluates whether prescription s has at least some medicines in the joint set of boxes j, j-1 and j+1; if the condition is met, then I(s,M[j]∪M[j-1]∪M[j+1])=1, otherwise it is 0.

[0028] Step (1) also includes integral fusion calculation: based on the calculated drug usage frequency F of the j-th box. jThe system yields four results: space utilization rate R[j], prescription correlation degree A[j], and prescription concentration degree between adjacent boxes C[j]. Users can set the importance of these four items through the system, with importance levels of w1 (important), w2 (relatively important), w3 (moderate), and w4 (not important). Setting w1 = 0.5, w2 = 0.3, w3 = 0.2, and w4 = 0, the system can then obtain w[j]. f w r w a w c Four parameters, here w f w r w a w c Each parameter (w1) represents a unique value from w1 to w4, and they do not repeat each other. In other words, each parameter (w1) represents a unique value from w1 to w4. f w r w a w c Let be a value from (w1, w2, w3, w4). Then, for box j, the integral can be calculated as:

[0029] T j =w f F j +w r R[j]+w a A[j]+w c C[j]

[0030] Based on the final ranking, the cell with the highest score T is selected as the box with the most suitable storage location.

[0031] Step (1) also includes: prescription entry: enter the prescription through the WIFI module provided by the medicine box and the mobile APP; the prescription includes the drug category, frequency of use, and dosage information, and the person to whom the prescription belongs should be specified when entering the prescription.

[0032] A home-use smart medicine cabinet based on a greedy strategy for storing and retrieving medicines includes the following modules:

[0033] The power supply module provides the 5V power required for system operation and maintains a stable voltage.

[0034] The facial recognition module is used for user authentication when picking up or storing medications;

[0035] The medicine box provides a WIFI module for remote connection and user interaction via WIFI, used to transmit data streams and receive analysis and processing results and control commands from the cloud server.

[0036] The scanning module is used to intuitively obtain relevant attributes of medicine boxes during the storage process, including size and drug information;

[0037] The storage module is used for data storage during the device's operation, including facial recognition results, prescription storage, and related process data.

[0038] The gyroscope control module is used to control the control switches of relevant medicine boxes in the medicine box, and controls the closing operation through algorithm scheduling results.

[0039] The voice module is used to remind users to take medication and give medication commands.

[0040] The beneficial effects of this invention are: Intelligent prescription integrated management: Traditional home medicine cabinets often only provide simple physical storage functions, while this algorithm achieves intelligent prescription management by considering the centralized storage of medicines in the prescription. It not only optimizes the physical layout of medicines but also improves the efficiency of quickly finding the required medicines in emergencies, reducing the risk of treatment delays due to searching for medicines.

[0041] Dynamic adaptive storage optimization: The algorithm dynamically adjusts the storage location based on the volume, shelf life, usage habits, and prescription requirements of medicines. Compared to a fixed-layout home medicine cabinet, this significantly improves space utilization and flexibility. It can automatically optimize storage strategies as family members' health conditions change and medicine inventory is updated, ensuring the medicine cabinet is always operating efficiently.

[0042] Personalized usage considerations: By incorporating the frequency of medication use into the algorithm, it can provide personalized storage suggestions based on the specific medication habits of family members. This addresses the issue of standard medicine cabinet products failing to adapt to diverse individual needs, thus enhancing the user experience.

[0043] Reducing medication waste: Prioritizing storage strategies with less remaining box space helps reduce wasted space. Good storage strategies indirectly promote the rational circulation and timely consumption of medications, preventing them from expiring due to improper storage, saving household expenses, and demonstrating environmental awareness. (When family members take frequently used medications, if similar medications that are about to expire are placed nearby, they are more likely to use those first, avoiding waste.)

[0044] Continuous optimization without frequent resets: The algorithm is designed to avoid redistributing existing medicines and instead select the optimal location for newly added medicines. This reduces the complexity and time cost of daily management, making medicine box management more convenient and sustainable.

[0045] Data analytics supports decision-making: The data processing and analysis capabilities behind the algorithms provide data support for family health management. Long-term application can collect data such as drug usage frequency and expiration date adherence, providing a scientific basis for future family medical resource planning and procurement plans. Attached Figure Description

[0046] Figure 1 A system structure diagram of a home-use smart medicine cabinet for storing and dispensing medicines.

[0047] Figure 2 A structural diagram illustrating the medicine storage and retrieval functions of a home-use smart medicine cabinet;

[0048] Figure 3 Flowchart for storing medicines in a home smart medicine cabinet;

[0049] Figure 4 A flowchart illustrating the medication dispensing process for a home smart medicine cabinet. Detailed Implementation

[0050] 1. Medicine box initialization

[0051] Define variables:

[0052] V[i] represents the volume of the i-th drug;

[0053] P[i] represents the shelf life of the i-th drug;

[0054] F[i] represents the frequency of use of the i-th drug;

[0055] B[j] represents the current remaining volume of the j-th box;

[0056] M[j] represents the set of medicines and their prescription information currently stored in the j-th box;

[0057] S[k] represents the set of medicines contained in the k-th prescription;

[0058] N is the known quantity of medicines;

[0059] S is the number of known prescriptions;

[0060] newDrug represents the drug information to be stored, including volume, prescription information, etc.

[0061] 2. Determining the correlation between prescriptions

[0062] For the drug to be stored (newDrug), query the prescription drug details table in the system to determine whether it is in the stored prescriptions.

[0063] First, perform a query operation in the prescription drug details table based on the name or identifier of newDrug. If the query returns the corresponding record, it means that newDrug already exists in the stored prescription, and steps 5 and 6 need to be performed.

[0064] 3. Consider centralized storage for high-frequency medication use.

[0065] For a new drug not listed in any prescription, sort all boxes that meet the required volume in descending order of usage frequency F[newDrug], and select the box with the highest frequency. First, query the drug usage history table for the past 180 days. If the new drug is in the historical drug usage list, count the number of days of usage L. days Number of uses L freq Total amount of medicine used (L) num Given that the recommended number of times drug i is taken per day is R. f And the recommended dosage R n Define the use of the drug frequency function:

[0066]

[0067] Here, w1, w2, and w3 are weighting coefficients, and w1 + w2 + w3 = 1. We set w1 = 0.5, w2 = 0.3, and w3 = 0.2.

[0068] For the i-th medicine box, iterate through all the medicines in the medicine box and calculate the total usage frequency F. j

[0069]

[0070] 4. Optimize space utilization

[0071] Based on the above conditions, for all possible boxes, select the one with the smallest remaining space that can still accommodate the new medicine, to reduce space waste. (Definable) To determine the space utilization of the current boxes, select the box with the largest R[j].

[0072] 5. Calculation of prescription correlation

[0073] Prescription affinity calculation aims to quantify how closely a newly added drug fits into the existing drug combinations in the medicine box to fulfill the prescription. We have found a method to measure which box a new drug should be stored in to best maintain or enhance the coverage of the current drug combinations in the medicine box. Below is a simplified prescription affinity calculation formula and its explanation:

[0074] Formula for calculating prescription correlation:

[0075]

[0076] Where: A[j] represents the prescription relevance score of box j; S is the set of all known prescriptions; N is the set of currently existing medicines in box j; i represents the i-th medicine, and k represents the k-th prescription; W k Indicates the degree of association between multiple drugs appearing in the same prescription; V i This represents the correlation between a drug appearing in multiple prescriptions; here, a function f(i,k) is defined such that if drug i exists in prescription k, then f(i,k) = 1, otherwise...

[0077] f(i,k)=0.

[0078] (1) Formula for calculating the correlation between multiple drugs appearing in the same prescription:

[0079]

[0080] G was used here. k The quadratic function as the prescription S k The weighted correlation reflects the higher the correlation bonus when more drugs appear in the same prescription.

[0081] (2) Formula for calculating the correlation of a drug appearing in multiple prescriptions

[0082]

[0083] H is used here. i The logarithm of i is used as the weighted correlation of drug i. The more prescriptions i appears in, the higher the correlation.

[0084] The explanation is as follows: Traversing all prescriptions: The formula first traverses all prescriptions s∈S to check if each prescription intersects with any medicine in the box under consideration, and to check the number of types of medicines with intersection. Traversing all medicines in all boxes: For each type of medicine, its frequency of appearance in all prescriptions is checked, i.e., to observe how many prescriptions require it. Cumulative scoring: By accumulating the matching scores of all prescriptions and the number of prescriptions containing all medicines, the overall prescription relevance score A[j] for box j is obtained. The higher the score, the more the combination of medicines in the box satisfies or nearly satisfies more prescription requirements, and the more the system tends to store multiple medicines from the same prescription in one box, thus awarding a higher score. Through the above formula and explanation, prescription relevance can be quantified, thereby aiding the decision-making process, ensuring that newly added medicines can be effectively integrated into the existing storage system, and promoting the rapid matching and efficient use of prescriptions.

[0085] 6. Formula for calculating the concentration of prescriptions in adjacent boxes

[0086] The calculation of prescription concentration in adjacent boxes aims to assess how well the medications stored in a given box and its adjacent boxes meet prescription needs. This helps maintain the physical proximity of related medications, allowing users to quickly find and use all medications in a prescription. Below is a formula for calculating prescription concentration in adjacent boxes and its explanation:

[0087] Formula for calculating the concentration of prescriptions in adjacent boxes:

[0088] Where: C[j] represents the prescription concentration score of box j and its adjacent boxes j-1 and j+1, assuming appropriate handling of boundary conditions (e.g., connected end to end or only considering one side of the neighbors of the boundary box); S is the set of all known prescriptions; M[j] is the set of medicines stored in box j, and M[j-1] and M[j+1] represent the sets of medicines in adjacent boxes. The specific adjacent boxes need to be calculated in practice.

[0089] I(s,M[j]∪M[j-1]∪M[j+1]) is an indicator function that evaluates whether at least some of the medicines in prescription s are present in the joint set of boxes j, j-1, and j+1. If the condition is met, then I(s,M[j]∪M[j-1]∪M[j+1])=log2(n), where n is the number of medicines in boxes j, j-1, and j+1 in the prescription; otherwise, it is 0.

[0090] Explanation: Expanded Scope of Consideration: Compared to considering only a single box's prescription relevance, this formula extends to consider the combined effect of the current box j and its left and right adjacent boxes (assuming a circular or linear sequence, and handling boundary cases). Extensiveness of Prescription Matching: For each prescription s, this formula checks whether the prescription can find enough ingredients in the current box j and its two adjacent boxes to satisfy the prescription. If all medicines or key medicine sets in at least one prescription can be found in these boxes, it indicates that the region is highly favorable for satisfying prescription needs. Cumulative Number of Satisfied Prescriptions: By summing the number of prescriptions that can be satisfied by the current box and its adjacent boxes, we obtain the C[j] value. The higher this value, the better the boxes in this region can support the rapid retrieval of prescriptions overall.

[0091] 7. Integral Fusion Calculation

[0092] Based on the calculations in steps 3, 4, 5, and 6, the overall drug usage frequency F of the j-th box was obtained. jThe system yields four results: space utilization rate R[j], prescription correlation degree A[j], and prescription concentration degree between adjacent boxes C[j]. Users can set the importance of these four items through the system, with importance levels of w1 (important), w2 (relatively important), w3 (moderate), and w4 (not important). Setting w1 = 0.5, w2 = 0.3, w3 = 0.2, and w4 = 0, the system can then obtain w[j]. f w r w a w c Four parameters, here w f w r w a w c Each parameter (w1) represents a unique value from w1 to w4, and they do not repeat each other. In other words, each parameter (w1) represents a unique value from w1 to w4. f w r w a w c Let be a value from (w1, w2, w3, w4). Then, for box j, the integral can be calculated as:

[0093] T j =w f F j +w r R[j]+w a A[j]+w c C[j]

[0094] Based on the final ranking, the cell with the highest score T is selected as the box with the most suitable storage location.

[0095] 9. Medication Collection Process

[0096] The diagram illustrates the workflow of a home smart medicine cabinet dispensing system. First, the user enters their identity information, including name and facial features, which the system then recognizes and records. Next, the user enters the prescription on a mobile app, including the recipient, medication name, dosage, and administration time. Based on this information, the system automatically associates the medications in the cabinet with the desired distribution. Then, it automatically schedules the medication dispensing time based on the entered prescription, providing an audio notification when it's time. Finally, after verifying their identity via facial recognition, the system searches the medication information table and prescription details table to determine the names and quantities of the medications in the prescription. It then checks the medication information table to determine if any medications are missing, and if so, provides a notification.

[0097] When the system detects that a user is picking up medication according to the prescription, it will select the appropriate box for the medication based on the following strategy: Since a medication may be placed in more than one box, to make it easier for the user to retrieve their medication, the system will pop up as few boxes as possible while still ensuring that the prescription contains the correct medication. Define the decision variable x. i (i = 1, 2, ..., m) are binary variables, where m is the number of cells containing the medicines in the prescription. If box i is selected, then x... i =1, otherwise x i =0, so that the objective function minimizes the number of boxes used. Furthermore, for each drug j (j = 1, 2, ..., n), a constraint is added to ensure that it is selected at least once. Where S i Let be the set of boxes containing medicines in the i-th box. By iterating through each medicine and checking which boxes contain that medicine, and ensuring that at least one of these boxes is selected, constraints are constructed to find as few boxes as possible that satisfy the prescription for medicines.

[0098] The system will then automatically dispense the medication and check if the medication collection is complete. If the user fails to collect the medication on time, the system will issue a timeout reminder. Once the user has collected the medication, the system will promptly update the medication information and remaining quantity to ensure the timeliness and accuracy of the information.

[0099] When users need to retrieve medication from the medicine cabinet, they can find the desired medication's compartment by name on the mobile app. After retrieving the medication and returning any remaining medication, they simply need to scan the barcode on the medication packaging again. The system will automatically update the remaining medication quantity based on the default dosage. If the medication is not returned, the system will invalidate the record for that medication. If it is not returned to the cabinet after a certain period, the record will be automatically cleared.

[0100] Data storage specifications

[0101] 1. Face recognition data structure

[0102] When designing a database to store facial features, it is necessary to consider the type of facial features and the way the data is organized. Facial recognition typically uses high-dimensional feature vectors for recording.

[0103]

[0104] 2. Drug Information Sheet

[0105] Drug ID INT A unique identifier for each drug Drug Name VARCHAR Full name of the medicine Production date DATE Drug production date Shelf life INT Shelf life days Instructions for use TEXT Instructions for use of medicine Box length DECIMAL Length of the packaging box (unit: millimeters) Box width DECIMAL Width of the packaging box (unit: millimeters) Box height DECIMAL Height of the packaging box (unit: millimeters) Warehouse Grid Number INT Warehouse cell number for storing this medicine Warehouse entry time DATETIME Time of drug storage

[0106] 3. Prescription Information Form

[0107] Prescription ID INT A unique identifier for each prescription number Prescription owner VARCHAR Information on the person to whom the prescription belongs Warehouse entry time DATETIME The time of prescription entry into the warehouse

[0108] 4. Prescription drug list

[0109]

[0110] 5. Storage Cell Details Table

[0111]

[0112] 6. Prescription Dispensing Record Form

[0113] Prescription ID INT A unique identifier for each prescription number Prescription owner VARCHAR Information on the person to whom the prescription belongs Drug dispensing time DATETIME The timing of the use of the medicine in this prescription .

Claims

1. A home-use intelligent medicine cabinet system for storing and retrieving medicines based on a greedy strategy, characterized in that: Includes a power supply module, a face recognition module, a WIFI module, a scanning module, a storage module, a servo control module, a voice module, a display module, and an embedded processor; includes the following: (1) Drug storage: acquire drug information, and then calculate the storage box based on the total usage frequency of the drugs, space utilization rate, prescription correlation degree, and prescription concentration of adjacent boxes; (2) Drug retrieval: users need to enter the identity information of the drug user, including name and facial feature information, and the system performs face recognition and records it; users enter prescriptions, including the prescription owner, drug name, dosage, and administration time information; the system calculates the drug based on the above. The system automatically associates information with the distribution of medications in the medicine box; then, based on the entered valid prescription, it automatically schedules the user's medication pickup time, and provides an audio prompt when it's time to take the medication; after the user verifies their identity, the system automatically matches the names and quantities of medications in the prescription based on the identified identity, and determines if any medications are missing, providing a prompt if so; when the system detects that the user is picking up medication according to the prescription, it selects the appropriate box for the medication according to the following strategy: since a medication may be placed in more than one box, to make it more convenient for the user to retrieve the medication, the system pops out as few boxes as possible while ensuring that the prescription contains all the medications; If newDrug is included in the history of drug use, the number of days of use should be recorded. Number of times used Total amount of medicine used Known drugs Recommended number of times per day And the recommended dosage per dose Define the use of the drug frequency function: ， Here, newDrug represents the drug information to be stored; , , It is a weighting coefficient, and ,set up , , For the first Given a medicine box, iterate through all the medicines in the box and calculate the total frequency of use. : ； Step (1) also includes: Space utilization optimization: For all possible boxes, select the box with the smallest remaining space that can still accommodate new medicines to reduce space waste; define R[j] = B[j] / Capacity[j] as the space utilization of the current box, and select the box with the smallest R[j]; where B[j] represents the current remaining volume of the j-th box; Step (1) also includes: Calculating prescription correlation degree: ， in: Indicates box The prescription relevance score; It is the collection of all known prescriptions; It's a box. The current collection of existing medicines; Indicates the first A type of medicine, Indicates the first Prescriptions for medicinal herbs; This indicates the degree of association between multiple drugs appearing in the same prescription. This represents the degree of association between a drug appearing in multiple prescriptions; a function is defined here. This makes if the medicine Existing in prescriptions Middle ,otherwise ; (1) Formula for calculating the correlation when multiple drugs appear in the same prescription: ， This uses The quadratic function as a prescription The weighted correlation reflects the higher the correlation bonus when more drugs appear in the same prescription. (2) Formula for calculating the correlation of a drug appearing in multiple prescriptions ， This uses The logarithm of a drug The weighted correlation coefficient increases the number of prescriptions in which the drug appears. Step (1) also includes: Formula for calculating the concentration of prescriptions in adjacent boxes: This aims to assess the extent to which the medicines stored in a box and its adjacent boxes meet prescription needs. This helps maintain the physical proximity of related medicines, making it easier for users to quickly find and use all the medicines in the entire prescription; Formula for calculating the concentration of prescriptions in adjacent boxes: ， in: Indicates box and its adjacent boxes and Assuming the boundary conditions are handled appropriately, the prescription concentration score of the first and last connected boxes or the boundary boxes only consider the prescription concentration score of one side of the neighbors. It is the collection of all known prescriptions; It's a box. The collection of drugs already stored in the database includes prescription information for these drugs. It is an indicator function that evaluates the prescription. Does at least some of the medicine exist in the box? , and In the union set; if the condition is satisfied, then Otherwise, it is 0; Step (1) also includes: integral fusion calculation: based on the calculated first... Total usage frequency of medicines in each box Space utilization rate Relevance of prescription Concentration of prescriptions in adjacent boxes The system assigns four importance levels to the four results. important, Relatively important generally, It's not important, and they won't overlap with each other, for the boxes. The integral is calculated as follows: ， Based on the final ranking, the cell with the highest score T is selected as the box with the most suitable storage location.

2. The home-use intelligent medicine cabinet system for storing and retrieving medicines based on a greedy strategy according to claim 1, characterized in that: The drug information includes: drug ID, drug name, production date, shelf life, instructions for use, box length, box width, and box height.

3. The home-use intelligent medicine cabinet system for storing and retrieving medicines based on a greedy strategy according to claim 1, characterized in that: Step (1) also includes: considering centralized storage of prescriptions: selecting target boxes: selecting boxes that meet the volume requirements and have the highest prescription relevance j* = argmax_j(A[j]), if there are multiple selections, then considering adjacency and utilization of remaining space; if there are multiple boxes that meet the conditions, further selection is based on the number of medicines with the same prescription in adjacent boxes, and a function can be defined. ,choose The biggest box.

Images