Medicine management method and system for old people

Abstract

The invention relates to the technical field of intelligent medicine management, in particular to a medicine management method and system for old people. Based on the medication data of the old people, the absorption, distribution, metabolism and excretion processes of the medicine in the body are simulated, data analysis is carried out, and medicine dynamic analysis auxiliary information is generated. According to the method, numerical simulation is carried out on the absorption rate of the medicine in the body, so that medicine management can accurately calculate the behavior and effect of the medicine in the body according to the individual difference of old people, and medicine dosage adjustment is provided for old patients, especially for the situation that the dosage needs to be accurately controlled, such as management of anticoagulant medicine. Through accurate simulation, the health risk caused by improper dosage of the medicine can be reduced to the greatest extent. Secondly, by constructing a state and reward matrix, data-driven decision support is provided for selection and dose adjustment of each medicine;

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent drug management, and in particular to a drug management method and system for the elderly. Background Art

[0002] The field of smart medication management technology is a field that combines information technology, the Internet of Things, artificial intelligence and other means to focus on optimizing and automating the medication management process. The development of the field aims to solve the problems existing in traditional medication management, such as errors in medication use, forgetting to take medication, and the risk of drug interactions, especially for the elderly population whose memory and executive function may decline. Through technologies such as smart devices (such as smart pill boxes, wearable devices), mobile applications, cloud computing and data analysis, smart medication management systems can remind users to take medication on time, monitor medication dosage, record medication use history, and provide warnings of drug interactions when necessary. The system can also enable medical providers to remotely monitor the medication intake of elderly patients, thereby providing more timely medical intervention.

[0003] A drug management method for the elderly is specifically targeted at a specific population of the elderly, and uses intelligent means to improve the efficiency and safety of drug management. Because the elderly may need to take multiple drugs at the same time due to cognitive decline, vision loss, or multiple diseases caused by aging, the difficulty of drug management and the risk of errors are greatly increased. The purpose of this method is to reduce risks and ensure that the elderly can use drugs correctly and safely, thereby improving their quality of life and health. By implementing this management method, the desired effects include reducing medication errors, improving medication compliance, reducing health risks caused by drug interactions, and supporting the elderly's ability to live independently through continuous drug management optimization.

[0004] Although traditional methods have achieved a certain degree of automation and monitoring in drug management, there is still a problem of insufficient response to dynamic changes in drugs. For example, in the management of medication for the elderly, it is difficult to accurately simulate the absorption and distribution process of drugs in the body, and it is not convenient to provide elderly patients with optimal drug dosage adjustments. In addition, traditional methods lack the ability to convert the trade-off between drug treatment effects and potential side effects into specific operational recommendations. For example, for elderly patients who suffer from multiple diseases and require multi-drug therapy, more sophisticated strategy adjustments are needed to find a balance between reducing the risk of cardiovascular events and avoiding excessive blood pressure reduction, and traditional methods are difficult to provide sufficient data to support the formulation of corresponding decisions. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a method and system for managing medicines for the elderly.

[0006] In order to achieve the above-mentioned object, the present invention adopts the following technical scheme: a method for managing medication for the elderly, comprising the following steps; S1: Based on the medication data of the elderly, simulate the absorption, distribution, metabolism and excretion process of drugs in the body, perform data analysis, and generate auxiliary information for drug dynamic analysis; S2: Based on the drug dynamic analysis auxiliary information, define the dynamic impact of drug concentration levels and health indicators of elderly patients, initialize the state and reward matrix, adjust the strategy according to the trade-off between treatment effect and side effect, and generate decision support auxiliary information; S3: Based on the decision support auxiliary information, simulate the arrival and processing process of the drug delivery request, record the waiting and service time of each request, optimize the human resource allocation according to the efficiency analysis, and generate process optimization suggestions; S4: Based on the process optimization suggestions, set drug inventory and dispensing costs as optimization targets, adjust inventory levels and distribution priorities, adjust dispensing quantities and schedules based on cost analysis, and generate cost-optimized resource allocation suggestions; S5: Based on the cost-optimized resource allocation suggestion, a drug distribution network is constructed, cost and time weights are assigned to each distribution path, a distribution route is selected according to cost efficiency, and a distribution path efficiency optimization suggestion is generated; S6: Based on the auxiliary information of drug dynamic analysis and combined with the drug taking history data of the elderly, data preprocessing and feature analysis are performed to train a drug demand prediction model and generate drug demand trend analysis results; S7: Based on the drug demand trend analysis results, combined with the real-time health data of the elderly, drug demand and health indicator data are integrated to analyze future drug demand, dynamically adjust drug inventory and distribution priorities according to the prediction results, and generate real-time dynamic adjustment strategies.

[0007] The present invention has the following improvements: the auxiliary information for drug dynamic analysis includes peak drug plasma concentration, drug half-life, and drug distribution ratio in tissues; the auxiliary information for decision support includes personal drug dosage range, recommended drug use time point, and interaction risk assessment of drug combination; the process optimization suggestion includes the number of service windows, key time nodes of drug review process, and response time target of consulting service; the resource allocation suggestion for cost optimization includes optimization suggestion for drug procurement batches, inventory turnover rate target, and layout from distribution center to service point; the distribution path efficiency optimization suggestion includes priority distribution path for emergency drugs, cost-effective route for conventional distribution, and distribution frequency adjustment; the drug demand trend analysis result includes seasonal drug demand change graph, drug demand forecast for disease outbreak period, and annual forecast for drug use trend of elderly population; the real-time dynamic adjustment strategy includes critical alarm value for drug inventory, key indicator threshold for health status monitoring, and medication adjustment reminder based on health data changes.

[0008] The present invention has been improved in that, based on the medication data of the elderly, the absorption, distribution, metabolism and excretion process of the drug in the body is simulated, and data analysis is performed to generate auxiliary information for drug dynamic analysis. The specific steps are as follows: S101: Based on the medication data of the elderly, perform numerical simulation of the drug absorption process, set the initial concentration of the drug in the stomach, adjust the time step, calculate the drug absorption rate, and generate the drug absorption rate analysis results; S102: Based on the drug absorption rate analysis result, simulate the distribution process of the drug in the body, set the drug distribution volume as an assumed value, perform integral solution, and adjust the integral accuracy to generate a drug distribution simulation result; S103: Based on the drug distribution simulation results, estimate the drug metabolism and excretion rates, input drug concentration and time data, set initial estimates of the metabolism rate and excretion rate, perform iterative optimization, and generate drug dynamic analysis auxiliary information.

[0009] The present invention is improved in that the drug absorption rate is calculated according to the improved formula I: ; Calculate the drug concentration at each time step; in, Indicates at time The drug concentration, Indicates at time The drug concentration, represents the absorption rate constant of the drug, represents the time step, Represents the renal function index of the elderly. Indicates the liver function index of the elderly. represents the pH adjustment factor of the drug, are the weights of renal function index, liver function index and pH adjustment coefficient respectively.

[0010] The present invention has the following improvements: based on the drug dynamic analysis auxiliary information, defining the dynamic impact of drug concentration levels and health indicators of elderly patients, initializing the state and reward matrix, adjusting the strategy according to the trade-off between treatment effect and side effect, and generating the decision support auxiliary information in the following specific steps; S201: Based on the auxiliary information of drug dynamic analysis, a state space including drug concentration levels and health indicators of elderly patients is defined using a Markov decision process, and a reward mechanism is constructed through action definitions such as drug selection and dosage adjustment to reflect the trade-off between treatment effect and side effects, thereby generating a state transition probability matrix; S202: Initializing a decision matrix based on the state transition probability matrix, and gradually updating the decision logic through balance analysis and utilization strategy to generate an optimized Q table; S203: Based on the optimized Q table, traverse and evaluate the decision effects under multiple states, select the optimal drug selection, medication time and dosage adjustment strategy under each state, and generate decision support auxiliary information.

[0011] The present invention is improved in that the Markov decision process is according to the improved formula II: ; Calculate the transition probability from the current health state to the next health state; in, Indicates from the state Through Action Transfer to state The probability of A description of a moment in time, including drug concentrations, patient health indicators such as heart rate or liver function index, Indicates the drug and dosage, is the probability function, Indicates the next time point status, Indicates the current time point status, Indicates at a point in time Actions taken, Indicates the heart rate change index of elderly patients. Indicates the interaction score between the drug and other drugs. represents the patient's liver function score, Represents the patient's quality of life score.

[0012] The present invention has the following improvements: based on the decision support auxiliary information, the arrival and processing of drug delivery requests are simulated, the waiting and service time of each request is recorded, the human resource allocation is optimized according to the efficiency analysis, and the specific steps of generating process optimization suggestions are as follows; S301: Based on the decision support auxiliary information, the drug delivery process is mapped, and an event-based simulation is performed to simulate the arrival and processing of drug delivery requests, and the waiting time and service time of each request are recorded to generate drug delivery process efficiency analysis results; S302: Based on the results of the drug distribution process efficiency analysis, perform a human resource configuration analysis for the drug review process, simulate the review capabilities under various configurations, calculate the average review completion time and queue length, optimize the review personnel configuration, and generate drug review process configuration recommendations; S303: Based on the drug review process configuration suggestion, combined with the instant consultation request data, the number of consultation service windows is adjusted, the customer waiting time and service efficiency under various numbers of service windows are simulated, and process optimization suggestions are generated.

[0013] The present invention has the following improvements: based on the process optimization suggestion, setting drug inventory and dispensing cost as optimization targets, adjusting inventory and delivery priority, adjusting dispensing quantity and time schedule according to cost analysis, and generating cost-optimized resource allocation suggestions in the following specific steps; S401: Based on the process optimization suggestion, balance the inventory holding cost and the ordering cost, set the drug ordering quantity, and generate an inventory cost optimization result; S402: Based on the inventory cost optimization result, the dispensing operation cost and the expired drug disposal cost are adjusted, the drug dispensing quantity and time schedule are set, the working hours of the dispensing staff and the drug expiration limit are reflected, the cost analysis is performed, and the dispensing process cost optimization suggestions are generated; S403: Based on the cost optimization suggestion for the dispensing process, set distribution priorities, control transportation and detention costs, reflect the demand satisfaction deadline and transportation resource constraints, select transportation routes, and conduct strategic analysis to generate cost-optimized resource allocation suggestions.

[0014] The present invention is improved in that, based on the cost-optimized resource allocation suggestion, a drug distribution network is constructed, cost and time weights are allocated to each distribution path, a distribution route is selected according to cost efficiency, and the specific steps of generating a distribution path efficiency optimization suggestion are as follows; S501: Based on the cost-optimized resource allocation suggestion, a drug distribution network model is constructed, distribution centers and receiving points are defined as network nodes, distribution paths are defined as network edges, cost and time weights are assigned to each edge, and a drug distribution network construction result is generated; S502: Based on the drug distribution network construction result, calculate the lowest cost path from the distribution center to the receiving point, refer to the carrying capacity of the route and the distribution time limit, perform path selection and calculation, and generate a distribution path selection result; S503: Based on the distribution path selection result, the distribution efficiency is evaluated, and the transportation process of the drugs from the distribution center to the receiving point is simulated, the transportation time and cost are recorded, the efficiency impact of the path selection is analyzed, and a distribution path efficiency optimization suggestion is generated.

[0015] The present invention is improved in that, based on the auxiliary information of drug dynamic analysis, combined with the drug taking history data of the elderly, data preprocessing and feature analysis are performed, a drug demand prediction model is trained, and the specific steps of generating drug demand trend analysis results are as follows; S601: Based on the auxiliary information of drug dynamic analysis and combined with the drug taking history data of the elderly, the missing values ​​in the data set are removed, data normalization is performed, and encoding conversion is performed on categorical variables to generate a preprocessed data set; S602: Based on the preprocessed data set, construct a decision tree model, control the number and depth of trees, train the model, evaluate its prediction performance, determine parameter configuration, and generate a drug demand prediction model; S603: Based on the drug demand prediction model, analyze the trend of future drug demand, including seasonal changes and abnormal points, draw a trend graph, and generate drug demand trend analysis results.

[0016] The present invention has the following improvements: based on the analysis results of the drug demand trend, combined with the real-time health data of the elderly, integrating drug demand and health index data, analyzing future drug demand, dynamically adjusting drug inventory and distribution priority according to the prediction results, and generating a real-time dynamic adjustment strategy; S701: Based on the drug demand trend analysis results, combined with the real-time health data of the elderly, perform data integration, merge the drug dynamic analysis information and health indicator data, set the timestamp as an index, and generate an integrated data set; S702: Based on the integrated data set, perform time series analysis, build a model to identify periodic changes and trends in demand, predict future drug demand, and generate a drug demand time series prediction model; S703: Based on the drug demand time series prediction model, health monitoring data is collected and analyzed in real time, drug inventory and distribution priority are adjusted according to the model output, and a real-time dynamic adjustment strategy is generated.

[0017] A drug management system for the elderly, the system comprising: The pharmacokinetic model analysis module is based on the medication data of the elderly. It analyzes the drug absorption, distribution, metabolism and excretion process through numerical simulation, sets the initial conditions and time steps, and generates auxiliary information for drug dynamic analysis. The decision optimization simulation module performs optimization simulation of drug management decisions based on the drug dynamic analysis auxiliary information, including drug selection, medication time and dosage adjustment, and generates decision support auxiliary information; The process optimization simulation module simulates the processes of drug distribution, review and consultation based on the decision support auxiliary information, optimizes waiting time and resource allocation, and generates process optimization suggestions; The cost optimization planning module designs a resource allocation plan for drug supply chain cost optimization based on the process optimization suggestions, including dynamic adjustment of inventory, dispensing and distribution priorities, and performs cost-benefit analysis to generate cost-optimized resource allocation suggestions; The path efficiency optimization module optimizes the drug distribution path and speed based on the cost-optimized resource allocation suggestion, evaluates the distribution efficiency, and generates a distribution path efficiency optimization suggestion; The demand prediction model module constructs a prediction model for future drug demand based on the drug dynamic analysis auxiliary information and the drug taking history data of the elderly, and performs demand trend analysis to generate drug demand trend analysis results; The real-time adjustment strategy module designs a dynamic adjustment strategy for drug inventory status and health indicators of elderly patients based on the drug demand trend analysis results and in combination with the real-time health data of the elderly, updates resource priorities, and generates a real-time dynamic adjustment strategy; The data integration module performs data integration operations based on the drug dynamic analysis auxiliary information, decision support auxiliary information, process optimization suggestions, cost-optimized resource allocation suggestions, distribution path efficiency optimization suggestions, drug demand trend analysis results and real-time dynamic adjustment strategies to generate a comprehensive data analysis set.

[0018] Compared with the prior art, the advantages and positive effects of the present invention are: In the present invention, by numerically simulating the absorption rate of drugs in the body, drug management can be based on individual differences in the elderly, so as to accurately calculate the behavior and effects of drugs in the body, and provide drug dosage adjustments for elderly patients, especially for those situations where the dosage needs to be precisely controlled, such as the management of anticoagulants. Through precise simulation, health risks caused by improper drug dosage can be minimized. Secondly, by constructing a state and reward matrix, data-driven decision support is provided for each drug selection and dosage adjustment. The method not only enhances the scientificity and adaptability of decision-making, but also can dynamically adjust the treatment plan to achieve the optimal trade-off between treatment effect and side effects. When dealing with elderly patients with multiple chronic diseases and multi-drug combination, the interaction risks of various drug combinations can be evaluated in real time, and the medication regimen can be adjusted to minimize side effects while maintaining the effectiveness of treatment. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 A flow chart of a method for drug management for the elderly is proposed for the present invention; Figure 2 The present invention proposes a detailed flow chart of step S1 in a method for managing medication for the elderly; Figure 3 The present invention proposes a detailed flow chart of step S2 in a method for managing medication for the elderly; Figure 4 The present invention proposes a detailed flow chart of step S3 in a method for managing medication for the elderly; Figure 5The present invention proposes a detailed flow chart of step S4 in a method for managing medication for the elderly; Figure 6 The present invention proposes a detailed flow chart of step S5 in a method for managing medication for the elderly; Figure 7 The present invention proposes a detailed flow chart of step S6 in a method for managing medication for the elderly; Figure 8 The present invention proposes a detailed flow chart of step S7 in a method for managing medication for the elderly; Fig. 9 The present invention proposes a module diagram of a drug management system for the elderly. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0021] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating positions or positional relationships, are based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.

[0022] For examples, see Figure 1 , the present invention provides a technical solution: a method for managing medication for the elderly, comprising the following steps; S1: Based on the medication data of the elderly, simulate the absorption, distribution, metabolism and excretion process of drugs in the body, perform data analysis, and generate auxiliary information for drug dynamic analysis; S2: Based on the auxiliary information of drug dynamic analysis, define the dynamic impact of drug concentration levels and health indicators of elderly patients, initialize the state and reward matrix, adjust the strategy according to the trade-off between treatment effect and side effect, and generate auxiliary information for decision support; S3: Based on decision support auxiliary information, simulate the arrival and processing of drug delivery requests, record the waiting and service time of each request, optimize human resource allocation based on efficiency analysis, and generate process optimization suggestions; S4: Based on the process optimization suggestions, set the drug inventory and dispensing cost as the optimization target, adjust the inventory and distribution priority, adjust the dispensing quantity and schedule according to the cost analysis, and generate cost-optimized resource allocation suggestions; S5: Based on cost-optimized resource allocation suggestions, a drug distribution network is constructed, cost and time weights are assigned to each distribution path, distribution routes are selected based on cost efficiency, and distribution path efficiency optimization suggestions are generated; S6: Based on the auxiliary information of drug dynamic analysis and combined with the historical data of drug taking of the elderly, data preprocessing and feature analysis are performed to train the drug demand prediction model and generate drug demand trend analysis results; S7: Based on the results of drug demand trend analysis, combined with the real-time health data of the elderly, drug demand and health indicator data are integrated to analyze future drug demand, dynamically adjust drug inventory and distribution priorities according to the forecast results, and generate real-time dynamic adjustment strategies.

[0023] Auxiliary information for drug dynamic analysis includes peak plasma concentration of drugs, drug half-life, and distribution ratio of drugs in tissues. Auxiliary information for decision support includes personal dosage range, recommended medication time points, and interaction risk assessment of drug combinations. Process optimization suggestions include the number of service windows, key time nodes in the drug review process, and response time targets for consulting services. Resource allocation suggestions for cost optimization include optimization suggestions for drug procurement batches, inventory turnover targets, and layout from distribution centers to service points. Distribution path efficiency optimization suggestions include priority distribution routes for emergency drugs, cost-effective routes for routine distribution, and distribution frequency adjustments. Drug demand trend analysis results include seasonal drug demand change charts, drug demand forecasts during disease outbreaks, and annual forecasts of drug use trends in the elderly population. Real-time dynamic adjustment strategies include critical alarm values ​​for drug inventory, key indicator thresholds for health status monitoring, and medication adjustment reminders based on changes in health data.

[0024] See also Figure 2 ,Based on the medication data of the elderly, the specific steps of simulating the absorption, distribution, metabolism and excretion process of drugs in the body, and conducting data analysis to generate auxiliary information for drug dynamic analysis are as follows; S101: Based on the medication data of the elderly, the specific process of numerically simulating the drug absorption process, setting the initial concentration of the drug in the stomach, adjusting the time step, calculating the drug absorption rate, and generating the drug absorption rate analysis results is as follows; Based on the medication data of the elderly, a numerical simulation of the drug absorption process is performed, the initial concentration of the drug in the stomach is set, and the time step is adjusted according to the characteristics of the drug and the physiological response of the elderly. The drug absorption rate is calculated using the Euler method. During the process, a simulation equation for the drug absorption rate is set, which describes the change in drug concentration in the stomach over time. Starting from the initial concentration, the drug concentration is gradually updated according to the time step, and this calculation is repeated until the set simulation time is completed. The drug concentration is repeatedly updated until the simulation ends, thereby generating concentration change data for the entire drug absorption process. In this process, the absorption rate constant is estimated based on previous experimental data. The absorption process of the drug in the body can be simulated and recorded, and a detailed chart of the drug concentration changing over time can be obtained.

[0025] Calculate the drug absorption rate according to the improved formula I: ; Calculate the drug concentration at each time step; in, Indicates at time of drug concentration. Indicates at time The drug concentration is an input variable in the formula and represents the drug concentration at the previous time step. The absorption rate constant of a drug is a coefficient that reflects the absorption rate of the drug and is determined based on the drug properties and the reaction rate in the body. Represents the time step, which is the length of time between each calculation during the simulation process, and is used to control the accuracy of the simulation and the speed of calculation. The renal function index of the elderly is a quantitative parameter that reflects the renal function status. It is obtained based on the health data of the elderly and is used to adjust the drug absorption rate to reflect the impact of renal function on drug processing capacity. The liver function index of the elderly is a parameter that quantifies the liver function status of the elderly. It is obtained based on health data and is used to adjust the drug absorption rate to reflect the impact of liver function on drug metabolism. It represents the pH adjustment factor of the drug, which is adjusted according to the stability and solubility of the drug under specific pH conditions and affects the absorption rate of the drug in the stomach. They are the weights of the renal function index, liver function index and pH adjustment coefficient respectively. The weights are determined through data analysis to reflect the influence of various physiological and chemical parameters on the drug absorption rate.

[0026] The execution process is as follows: Suppose there is a set of health data ,in and Represents the real-time physiological indexes of renal function and liver function respectively. The indexes are converted into a unified quantitative index through a standardized method. For example, using the linear conversion formula and ,in It is a coefficient obtained through historical data analysis.

[0027] Solubility data of drugs at different pH values Can be obtained through testing.

[0028] Assume that the current pH of gastric acid in elderly patients is ,but Can be based on interpolation or regression models To predict the current drug solubility state.

[0029] Through multivariate regression analysis of historical data, the relative importance of each factor on drug absorption is determined, and the weight is then set. .

[0030] First, set the initial drug concentration and time step .

[0031] calculate And weight .

[0032] For each time point ,calculate Using the improved formula I: , renew for , and repeat the calculation until the entire simulation period is completed.

[0033] The obtained drug concentration change diagram It can more accurately reflect the drug absorption process of the elderly, taking into account their renal function, liver function and drug chemical properties (such as pH sensitivity). This makes the adjustment of drug dosage more personalized, enhancing the safety and effectiveness of treatment, which is especially important in clinical scenarios that require precise control of drug concentrations.

[0034] S102: Based on the drug absorption rate analysis results, simulate the distribution process of the drug in the body, set the drug distribution volume as an assumed value, perform integral solution, and adjust the integral accuracy. The specific process of generating the drug distribution simulation result is as follows; Based on the results of drug absorption rate analysis, simulate the distribution process of drugs in the body, set the drug distribution volume as the assumed value Vd, perform integral solution, and adjust the integral precision to generate drug distribution simulation results. Using the results of drug absorption as input, set the dynamic model of drug distribution in the body as Vd*dC / dt=-k_d*C, where k_d represents the rate constant of drug distribution, and use the numerical integration method to calculate the overall distribution of drugs within time t. Specifically, by setting the initial value C0 and updating the drug concentration C at each time step Δt, calculate the drug concentration in the body at each moment. The accuracy of the simulation is ensured by adjusting the step size and precision of the integration, so as to obtain the overall distribution of the drug in the body.

[0035] S103: Based on the drug distribution simulation results, the drug metabolism and excretion rates are estimated, drug concentration and time data are input, initial estimates of the metabolism rate and excretion rate are set, and iterative optimization is performed to generate the specific process of auxiliary information for drug dynamic analysis: Based on the results of drug distribution simulation, the drug metabolism and excretion rates are estimated, the drug concentration and time data are input, the initial estimated values ​​of the metabolic rate and excretion rate k_m and k_e are set, iterative optimization is performed, and auxiliary information for drug dynamic analysis is generated. Here, the model of drug metabolism and excretion is constructed as dC / dt=-(k_m+k_e)*C, where k_m and k_e are the rate constants of metabolism and excretion, respectively. Through the input drug concentration C and time t data points, the values ​​of k_m and k_e are optimized by iterative method to minimize the difference between the predicted and actual concentrations. During the optimization process, by adjusting k_m and k_e, iterative solution is obtained until the convergence condition or the maximum number of iterations is reached, thereby obtaining the final estimated values ​​of metabolism and excretion rates. The adjustment and optimization of parameters are based on actual drug dynamic observation data and the physiological characteristics of the elderly population.

[0036] See also Figure 3 ,Based on the auxiliary information of drug dynamic analysis, the dynamic ,impact of drug concentration levels and health indicators of elderly patients ,is defined, the state and reward matrix are initialized, the ,strategy is adjusted according to the trade-off between treatment ,effects and side effects, and the specific steps of generating ,decision support auxiliary information are as follows; S201: Based on the auxiliary information of drug dynamic analysis, the state space including drug concentration levels and health indicators of elderly patients is defined using Markov decision process. Through action definitions such as drug selection and dosage adjustment, a reward mechanism is constructed to reflect the trade-off between treatment effect and side effects. The specific process of generating the state transition probability matrix is ​​as follows; Based on the auxiliary information of drug dynamic analysis, the Markov decision process is used to define the state space covering drug concentration and health indicators of elderly patients, and the corresponding action space is set to include different drug types and dosage combinations. The combination is used to form specific treatment decisions. A reward mechanism is further constructed to balance the treatment effect with possible side effects to guide the best treatment action selection. In actual operation, the state space consists of the concentration of drugs in the body and the key indicators affecting the health of elderly patients, while the action space consists of the available drugs and their doses. The reward mechanism is set to evaluate the treatment effect and side effects of each decision. By weighing the factors, the reward mechanism helps determine the most appropriate treatment plan. In addition, the probability of transitioning from one health state to another needs to be calculated, which depends on the current health state, the selected treatment method and its dose. The calculation of the transition probability is based on the existing drug dynamic data and historical treatment records. Finally, a comprehensive Markov decision process model can be established, which can provide in-depth analysis and support for drug treatment decisions. Make the medical decision-making process more data-driven and result-oriented, aiming to optimize the treatment effect of elderly patients and minimize side effects.

[0037] The Markov decision process follows the improved formula II: ; Calculate the transition probability from the current health state to the next health state; in, Indicates from the state Through Action Transfer to state The probability of . It is used to describe the possibility of transitioning to the next state given the current state and action. It refers to the description at a certain moment, including the concentration of the drug, the patient's health indicators such as heart rate or liver function index, etc. Denotes medication and dosage. And interventions taken in a given state, aimed at changing the state or maintaining a certain state, such as adjusting medication dosage or changing treatment regimen. It is a probability function, which is used to calculate the probability of an event occurring under certain conditions. Indicates the next time point The state is the result state of probability calculation. Indicates the current time point status is the basis for decision making. Indicates at a point in time The actions taken depend on different treatment strategies such as drug selection and dose adjustment. It represents the heart rate change index of elderly patients, which is used to reflect the patient's heart health status and affect the effect of drug treatment. It indicates the interaction score between a drug and other drugs, which is used to evaluate the mutual effects and potential risks when the drugs are used together. It indicates the patient's liver function score, which critically affects the metabolic process and rate of drugs and has a direct impact on drug processing capacity. It represents the patient's quality of life score, assesses the overall impact of treatment on the patient's daily life, and is an important factor in considering the trade-off between treatment effect and side effects.

[0038] The execution process is as follows: Heart rate data of the elderly patient are collected regularly, for example three times a day.

[0039] Use moving average or exponential smoothing methods to stabilize heart rate data and reduce the impact of occasional fluctuations.

[0040] The processed heart rate data is converted into a standard score, namely , where h is the smoothed heart rate data, and are the mean and standard deviation of historical heart rate data respectively.

[0041] Potential drug interactions were evaluated using the drug interaction database according to the prescribing information for elderly patients.

[0042] Based on clinical study data and pharmacological information, each potential drug combination is assigned a risk score.

[0043] Calculate the weighted average interaction score for the current drug combination , where each individual score is weighted by the severity of its impact.

[0044] Regularly check liver function test results, such as ALT and AST levels.

[0045] Liver function score is calculated based on liver function test results using an established clinical score , including algorithms and tables, to quantify liver damage or reduced function.

[0046] Data were collected using standardized quality-of-life questionnaires, such as the SF-36.

[0047] The total score was extracted from the questionnaire results and standardized to obtain the quality of life score .

[0048] Using the above parameters Adjust the basic state transition probability.

[0049] The specific calculation is: ; in, is the basic transition probability without considering additional parameters, It is the weight of each parameter. The weight is determined through data analysis and reflects the influence of each factor on the transfer probability.

[0050] The resulting state transition probability matrix not only reflects the dynamic changes based on drug and health status, but also integrates factors such as the patient's physiological response and quality of life. This enables the model to more comprehensively and accurately predict health transitions under different treatment options, thereby helping medical providers make more reasonable drug treatment decisions, optimize treatment outcomes, reduce side effects, and improve patient satisfaction and quality of life.

[0051] S202: Based on the state transition probability matrix, the decision matrix is ​​initialized, and the decision logic is gradually updated through balance analysis and utilization strategy to generate the optimized Q table. The specific process is as follows; Based on the state transition probability matrix, the decision matrix is ​​initialized, and the decision logic is gradually updated through balance analysis and utilization strategies to generate an optimized Q table. The Q value in all states is initially set to zero, that is, Q(s, a)=0. During the simulation, the ε-greedy strategy is used to select actions, that is, the action with the highest Q value in the current state is selected with a probability of 1-ε, and any action is randomly selected with a probability of ε to explore the new state space. After each action, the Q value is updated according to the observed reward and new state.

[0052] S203: Based on the optimized Q table, traverse and evaluate the decision effects under multiple states, select the optimal drug selection, medication time and dosage adjustment strategy under each state, and generate the specific process of decision support auxiliary information; Based on the optimized Q table, the decision effects under multiple states are traversed and evaluated, and the optimal drug selection, medication time and dosage adjustment strategy under each state are selected to generate decision support auxiliary information. The traversal operation includes finding the action a that maximizes the Q value in each state s, that is, a*=argmax_aQ(s, a). Ensure that each state selects a strategy that can maximize the expected long-term return. And provide the recommended drug type, dosage and time of administration under each specific state. Decision support information can be directly applied to personalized drug management for elderly patients to improve the individualized accuracy and safety of treatment.

[0053] See also Figure 4 ,Based on the decision support auxiliary information, the arrival and processing of ,drug delivery requests are simulated, the waiting and service time of each ,request is recorded, and the human resource allocation is optimized according to ,efficiency analysis. The specific steps of generating process optimization ,suggestions are as follows; S301: Based on the decision support auxiliary information, the drug distribution process is mapped, and an event-based simulation is performed to simulate the arrival and processing of drug distribution requests, and the waiting time and service time of each request are recorded to generate the specific process of the drug distribution process efficiency analysis result; Based on decision support auxiliary information, the drug distribution process is mapped, and event-based simulation is performed to simulate the arrival and processing of drug distribution requests, record the waiting time and service time of each request, and generate the results of drug distribution process efficiency analysis. In this process, the various stages of drug distribution are first defined, including order reception, drug preparation, review, and distribution. Standard operation time is set for each stage, and the arrival of requests is simulated through the event simulation model. The arrival time and processing time are randomly generated, and the time is set based on the probability distribution of historical data. During the simulation, the waiting and processing time of each request is recorded, and the data is used to calculate the average waiting time and service time of the overall process, so as to evaluate the bottleneck and efficiency in the process.

[0054] S302: Based on the results of the drug distribution process efficiency analysis, the human resource configuration analysis of the drug review process is performed, the review capabilities under various configurations are simulated, the average review completion time and queue length are calculated, the review personnel configuration is optimized, and the specific process of generating drug review process configuration suggestions is as follows; Based on the results of the drug distribution process efficiency analysis, the human resource configuration analysis of the drug review link is carried out, the review capabilities under various configurations are simulated, the average review completion time and queue length are calculated, the review staff configuration is optimized, and the drug review process configuration suggestions are generated. This link focuses on optimizing human resources to improve review efficiency. The analysis method includes setting different staffing schemes, simulating the review process under each scheme, and recording the time required to complete the review and the queue length generated. By comparing the review efficiency under different configurations, the optimal staffing configuration is identified. Using a linear programming model, minimizing the review time and the shortest queue length are set as the goals, and the entire review process is optimized by adjusting the number of reviewers and allocation strategies.

[0055] S303: Based on the drug review process configuration suggestion, combined with the instant consultation request data, the number of consultation service windows is adjusted, and the customer waiting time and service efficiency under various numbers of service windows are simulated. The specific process of generating process optimization suggestions is as follows; Based on the configuration suggestions of the drug review process, combined with the instant consultation request data, the number of consultation service windows is adjusted, the customer waiting time and service efficiency under various numbers of service windows are simulated, and process optimization suggestions are generated. First, the historical data of consultation service requests are analyzed to determine the peak hours and service demand, and then different service window configuration simulation schemes are set. Through simulation analysis of customer waiting time and service window utilization under each scheme, statistical analysis methods are used to evaluate the service efficiency under different configurations. The results will guide the adjustment of the number of consultation windows to balance customer waiting time and service cost and achieve the optimization of service process.

[0056] See also Figure 5,Based on the process optimization suggestions, setting drug inventory and dispensing cost as optimization targets,adjusting inventory and distribution priorities, adjusting dispensing quantity and schedule according to cost analysis,,the specific steps for generating cost-optimized resource allocation suggestions are as follows; S401: Based on the process optimization suggestions, balance the inventory holding cost and ordering cost, set the drug ordering quantity, and generate the specific process of inventory cost optimization results; Based on the process optimization suggestions, balance the inventory holding cost and ordering cost, set the drug ordering quantity, and generate inventory cost optimization results. First, analyze the frequency of use and expiration risk of the drug to determine the ideal inventory level for each drug. The economic ordering batch model is used to determine the number of drugs ordered each time. The model takes into account the trade-off between ordering cost and holding cost to minimize the total cost. Specifically, the ordering cost and holding cost functions are set, and the differential solution method is used to find the drug ordering quantity that minimizes the total cost. It can effectively reduce the situation of overstocking or out-of-stock, ensuring drug supply and cost efficiency.

[0057] S402: Based on the inventory cost optimization results, the specific process of adjusting the dispensing operation cost and the expired drug disposal cost, setting the drug dispensing quantity and time schedule, reflecting the working hours of the dispensing staff and the drug expiration limit, performing cost analysis, and generating the dispensing process cost optimization suggestions is as follows; Based on the inventory cost optimization results, adjust the dispensing operation cost and expired drug disposal cost, set the drug dispensing quantity and schedule, reflect the dispensing staff's working hours and drug expiration limit, perform cost analysis, and generate dispensing process cost optimization suggestions. This link optimizes costs by analyzing the time cost of dispensing operations and the cost of drug disposal in detail. First, adjust the dispensing quantity according to the drug's expiration date and demand forecast to avoid drug expiration. Then, reduce human resource waste by optimizing the dispensing staff's work schedule and process layout. Use time motion study methods to evaluate and optimize each step of the dispensing operation to ensure that each operation is run at the lowest cost.

[0058] S403: Based on the cost optimization suggestions for the dispensing process, the specific process of setting distribution priorities, controlling transportation and detention costs, reflecting the demand satisfaction deadline and transportation resource constraints, selecting transportation routes, and conducting strategy analysis to generate cost-optimized resource allocation suggestions is as follows; Based on the cost optimization suggestions for the dispensing process, set distribution priorities, control transportation and detention costs, reflect the demand fulfillment deadline and transportation resource constraints, select transportation routes, and conduct strategic analysis to generate cost-optimized resource allocation suggestions. Set different distribution priorities according to the urgency and shelf life of the drug, and use cost-benefit analysis to select the most economical transportation route. Use linear programming models to allocate transportation resources, optimize route selection and transportation scheduling, and reduce transportation and warehousing costs. It can minimize logistics costs while ensuring that the drugs arrive on time.

[0059] See also Figure 6 ,Based on the cost-optimized resource allocation proposal, the drug distribution network is constructed,,cost and time weights are assigned to each distribution path, and the,delivery route is selected based on cost efficiency.,The specific steps for generating the distribution path efficiency,optimization proposal are as follows; S501: Based on the cost-optimized resource allocation suggestion, a drug distribution network model is constructed, distribution centers and receiving points are defined as network nodes, distribution paths are defined as network edges, cost and time weights are assigned to each edge, and the specific process of generating the drug distribution network construction result is as follows; Based on the resource allocation suggestion of cost optimization, a drug distribution network model is constructed, distribution centers and receiving points are defined as network nodes, distribution paths are defined as network edges, cost and time weights are assigned to each edge, and the drug distribution network construction results are generated. First, it is clear that the nodes in the network include drug warehouses, hospitals, pharmacies and other receiving points. The connection path between each node represents a possible drug transmission route, and the weight of the path consists of two parts: transportation cost and time. The transportation cost is calculated based on the distance, transportation method (such as land transportation, air transportation) and related expenses; the time weight takes into account factors such as the length of the journey and traffic conditions. Use graph theory algorithms to weave these nodes and edges into a comprehensive network model to facilitate subsequent path optimization analysis.

[0060] S502: Based on the drug distribution network construction result, the minimum cost path from the distribution center to the receiving point is calculated, and the path selection and calculation are performed with reference to the carrying capacity and distribution time limit of the route. The specific process of generating the distribution path selection result is as follows; Based on the results of the drug distribution network construction, the lowest cost path from the distribution center to the receiving point is calculated. The path selection and calculation are performed with reference to the carrying capacity and distribution time limit of the route to generate the distribution path selection result. The Dijkstra algorithm is used to find the path with the lowest cost. The algorithm optimizes the shortest path from the starting point to each receiving point in the entire network by gradually confirming the shortest path points. The selection of each path also needs to consider the urgency and shelf life limit of the drug to ensure that all drugs can reach the designated location in time within the validity period. The total cost and time required for each possible path are automatically adjusted and calculated through software tools to determine the optimal path.

[0061] S503: Based on the distribution path selection result, the distribution efficiency is evaluated, the transportation process of the drug from the distribution center to the receiving point is simulated, the transportation time and cost are recorded, the efficiency impact of the path selection is analyzed, and the specific process of generating distribution path efficiency optimization suggestions is as follows; Based on the distribution path selection results, the distribution efficiency is evaluated. The transportation process of drugs from the distribution center to the receiving point is simulated, the transportation time and cost are recorded, the efficiency impact of the path selection is analyzed, and the distribution path efficiency optimization suggestions are generated. In this stage, the actual drug transportation process is simulated using logistics simulation software, and the transportation time and related costs of each path are recorded and analyzed. The simulation takes into account various variables, such as traffic delays, transportation interruptions, etc., to ensure the practicality of the results. According to the simulation results, the efficiency and cost-effectiveness of each path are evaluated, and the deficiencies are adjusted. Finally, specific optimization measures are proposed, such as increasing the transportation frequency of certain routes, adjusting some routes, or replacing more efficient transportation methods, to improve the efficiency and cost-effectiveness of the overall distribution network.

[0062] See also Figure 7 ,Based on the auxiliary information of drug dynamic analysis, combined with the historical data of drug taking,of the elderly, data preprocessing and feature analysis are performed,,the drug demand prediction model is trained, and the specific steps of generating drug demand trend,analysis results are as follows; S601: Based on the auxiliary information of drug dynamic analysis, combined with the drug taking history data of the elderly, the missing values ​​in the data set are removed, the data is normalized, and the encoding of the categorical variables is converted to generate the specific process of the preprocessed data set; Based on the auxiliary information of drug dynamic analysis and combined with the historical data of drug use of the elderly, the missing values ​​in the data set are removed, the data is normalized, and the categorical variables are encoded and converted to generate a preprocessed data set. In this step, data cleaning technology is first applied to identify and remove missing or outliers to ensure data quality. For continuous data, standardization is performed to make its mean 0 and standard deviation 1 to eliminate the influence of data of different magnitudes. For categorical variables such as drug type and frequency of use, the unique hot encoding method is used to convert them into numerical values ​​to make them suitable for subsequent model training. The consistency and applicability of the data set are ensured, laying a solid foundation for building an efficient prediction model.

[0063] S602: Based on the preprocessed data set, a decision tree model is constructed, the number and depth of trees are controlled, the model is trained, and its prediction performance is evaluated, parameter configuration is determined, and the specific process of generating a drug demand prediction model is as follows; Based on the preprocessed data set, a decision tree model is constructed, the number and depth of the tree are controlled, the model is trained and its prediction performance is evaluated, the parameter configuration is determined, and a drug demand prediction model is generated. Using the decision tree algorithm, the appropriate tree depth and number are set to avoid overfitting while maintaining the generalization ability of the model. The cross-validation method is used to select the optimal parameter configuration, such as the segmentation criterion and the minimum number of leaf node samples, and the decision tree structure is optimized. After the model training is completed, the test data set is used to evaluate its prediction performance, and the model performance is quantified by indicators such as accuracy, recall rate, and F1 score. The decision tree model that can accurately predict the drug demand of the elderly provides a basis for drug inventory management and distribution optimization.

[0064] S603: Based on the drug demand prediction model, the trend of future drug demand is analyzed, including seasonal changes and abnormal points, and a trend chart is drawn. The specific process of generating drug demand trend analysis results is as follows; Based on the drug demand forecasting model, analyze the trend of future drug demand, including seasonal changes and abnormal points, draw trend charts, and generate drug demand trend analysis results. Use the forecasting model to analyze the time series data of drug demand and identify seasonal patterns and potential demand peaks or troughs. Through statistical analysis methods such as time series decomposition, clarify the periodicity and abnormal fluctuations of demand changes. Draw trend charts and demand forecast charts to intuitively display drug use trends and predict future demand changes. Help medical institutions and drug suppliers optimize inventory management, ensure the balance between drug supply and elderly patients' needs, and improve service efficiency and elderly patients' satisfaction.

[0065] See also Figure 8 ,Based on the results of drug demand trend analysis, combined with the real-time health data of the elderly, ,the drug demand and health indicator data are integrated to analyze future drug demand, ,and dynamically adjust drug inventory and distribution priorities according to the prediction results, ,and the specific steps for generating real-time dynamic adjustment strategies are as follows; S701: Based on the results of the drug demand trend analysis, combined with the real-time health data of the elderly, data integration is performed to merge the drug dynamic analysis information and health indicator data, and the timestamp is set as the index. The specific process of generating the integrated data set is as follows; Based on the results of drug demand trend analysis, combined with the real-time health data of the elderly, data integration is performed to merge drug dynamic analysis information and health indicator data, set the timestamp as the index, and generate an integrated data set. In this step, the drug demand data is combined with the real-time health monitoring data of elderly patients, such as heart rate, blood pressure and other indicators, to form a multi-dimensional data set. Set a unified timestamp index for each data item to ensure time alignment for subsequent time series analysis. Through SQL query and data processing functions, clean and integrate data from different sources to ensure data consistency and integrity, and provide support for accurate demand forecasting and drug distribution.

[0066] S702: Based on the integrated data set, time series analysis is performed to build a model to identify the cyclical changes and trends in demand, predict future drug demand, and generate a drug demand time series prediction model. The specific process is as follows; Based on the integrated data set, time series analysis is performed to build a model to identify cyclical changes and trends in demand, predict future drug demand, and generate a drug demand time series prediction model. Use time series analysis techniques in statistics, such as the autoregressive integrated moving average model or the seasonal decomposition time series prediction model, to process and analyze the data. In the process of model building, seasonal adjustment and trend decomposition of the data are first performed, and then the model parameters, such as lag terms, difference orders, and moving average terms, are determined for model training and verification. Through diagnostic testing of the model, such as residual analysis, the accuracy and applicability of the model are ensured, providing a basis for effectively predicting future drug demand.

[0067] S703: Based on the drug demand time series prediction model, real-time collection and analysis of health monitoring data, adjustment of drug inventory and distribution priority according to the model output, and generation of a real-time dynamic adjustment strategy are the specific processes; Based on the drug demand time series prediction model, health monitoring data is collected and analyzed in real time, and drug inventory and distribution priorities are adjusted according to the model output to generate real-time dynamic adjustment strategies. By continuously collecting health monitoring data of the elderly and updating the input of the time series prediction model in real time. According to the model prediction results, the inventory and distribution plan of the drugs are dynamically adjusted to give priority to drug distribution in high-demand areas. In addition, inventory and distribution strategies can be adjusted quickly in response to emergencies such as flu season or other health crises. This ensures the timeliness and efficiency of drug supply, reduces the risk of drug shortages or surpluses, and improves the quality and response speed of overall medical services.

[0068] See also Fig. 9 , a drug management system for the elderly, the system comprising: The pharmacokinetic model analysis module is based on the medication data of the elderly. It analyzes the drug absorption, distribution, metabolism and excretion process through numerical simulation, sets the initial conditions and time steps, and generates auxiliary information for drug dynamic analysis. The decision optimization simulation module performs optimization simulation of drug management decisions based on drug dynamic analysis auxiliary information, including drug selection, medication time and dosage adjustment, and generates decision support auxiliary information; The process optimization simulation module simulates the processes of drug distribution, review and consultation based on decision support auxiliary information, optimizes waiting time and resource allocation, and generates process optimization suggestions; The cost optimization planning module designs a resource allocation plan for drug supply chain cost optimization based on process optimization suggestions, including dynamic adjustment of inventory, dispensing and distribution priorities, and conducts cost-benefit analysis to generate cost-optimized resource allocation suggestions; The path efficiency optimization module optimizes the drug distribution path and speed based on cost-optimized resource allocation suggestions, evaluates distribution efficiency, and generates distribution path efficiency optimization suggestions; The demand forecasting model module builds a forecasting model for future drug demand based on auxiliary information of drug dynamic analysis and historical data of drug use by the elderly, and conducts demand trend analysis to generate drug demand trend analysis results; The real-time adjustment strategy module designs a dynamic adjustment strategy based on the drug demand trend analysis results and the real-time health data of the elderly, and updates the resource priority to generate a real-time dynamic adjustment strategy. The data integration module performs data integration operations and generates a comprehensive data analysis set based on auxiliary information of drug dynamic analysis, auxiliary information of decision support, process optimization suggestions, resource allocation suggestions for cost optimization, distribution path efficiency optimization suggestions, drug demand trend analysis results and real-time dynamic adjustment strategies.

[0069] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A method for managing medication for the elderly, characterized in that: The steps include: Based on the medication data of the elderly, the specific steps of simulating the absorption, distribution, metabolism and excretion process of drugs in the body, and performing data analysis to generate auxiliary information for drug dynamic analysis are as follows; Based on the medication data of the elderly, the drug absorption process is numerically simulated, the initial concentration of the drug in the stomach is set, the time step is adjusted, the drug absorption rate is calculated, and the drug absorption rate analysis results are generated; Based on the drug absorption rate analysis result, simulate the distribution process of the drug in the body, set the drug distribution volume as an assumed value, perform integral solution, and adjust the integral accuracy to generate a drug distribution simulation result; Based on the drug distribution simulation results, estimate the drug metabolism and excretion rates, input drug concentration and time data, set initial estimates of the metabolism rate and excretion rate, perform iterative optimization, and generate drug dynamic analysis auxiliary information; The drug absorption rate is calculated according to the improved formula I: ； Calculate the drug concentration at each time step; in, Indicates at time The drug concentration, Indicates at time The drug concentration, represents the absorption rate constant of the drug, represents the time step, Represents the renal function index of the elderly. Indicates the liver function index of the elderly. represents the pH adjustment factor of the drug, are the weights of renal function index, liver function index and pH adjustment coefficient, respectively; Based on the drug dynamic analysis auxiliary information, the dynamic impact of drug concentration levels and health indicators of elderly patients is defined, the state and reward matrix are initialized, the strategy is adjusted according to the trade-off between treatment effect and side effect, and decision support auxiliary information is generated; Based on the drug dynamic analysis auxiliary information, the dynamic impact of drug concentration levels and health indicators of elderly patients is defined, the state and reward matrix are initialized, and the strategy is adjusted according to the trade-off between treatment effect and side effect. The specific steps of generating decision support auxiliary information are as follows; Based on the auxiliary information of drug dynamic analysis, a state space including drug concentration levels and health indicators of elderly patients is defined using a Markov decision process, and a reward mechanism is constructed through action definitions such as drug selection and dosage adjustment to reflect the trade-off between therapeutic effects and side effects, thereby generating a state transition probability matrix; Based on the state transition probability matrix, the decision matrix is ​​initialized, and the decision logic is gradually updated through balance analysis and utilization strategy to generate an optimized Q table; Based on the optimized Q table, traverse and evaluate the decision effects under multiple states, select the optimal drug selection, medication time and dosage adjustment strategy under each state, and generate decision support auxiliary information; The Markov decision process is based on the improved formula II: ； Calculate the transition probability from the current health state to the next health state; in, Indicates from the state Through Action Transfer to state The probability of A description of a moment in time, including drug concentrations, patient health indicators such as heart rate or liver function index, Indicates the drug and dosage, is the probability function, Indicates the next time point status, Indicates the current time point status, Indicates at a point in time Actions taken, Indicates the heart rate change index of elderly patients. Indicates the interaction score between the drug and other drugs. represents the patient's liver function score, represents the patient's quality of life score; Based on the decision support auxiliary information, simulate the arrival and processing process of drug delivery requests, record the waiting and service time of each request, optimize human resource allocation according to efficiency analysis, and generate process optimization suggestions; Based on the process optimization suggestions, set drug inventory and dispensing costs as optimization targets, adjust inventory levels and delivery priorities, adjust dispensing quantities and schedules based on cost analysis, and generate cost-optimized resource allocation suggestions; Based on the cost-optimized resource allocation suggestions, a drug distribution network is constructed, cost and time weights are assigned to each distribution path, a delivery route is selected according to cost efficiency, and a distribution path efficiency optimization suggestion is generated; Based on the auxiliary information of drug dynamic analysis, combined with the drug taking history data of the elderly, data preprocessing and feature analysis are performed to train a drug demand prediction model and generate drug demand trend analysis results; Based on the results of the drug demand trend analysis, combined with the real-time health data of the elderly, drug demand and health indicator data are integrated to analyze future drug demand, dynamically adjust drug inventory and distribution priorities according to the prediction results, and generate real-time dynamic adjustment strategies.

2. The method for managing medication for the elderly according to claim 1, characterized in that: The drug dynamic analysis auxiliary information includes the peak plasma concentration of the drug, the half-life of the drug, and the distribution ratio of the drug in the tissue. The decision support auxiliary information includes the personal medication dosage range, the recommended medication time point, and the interaction risk assessment of the drug combination. The process optimization suggestions include the number of service windows, the key time nodes of the drug review process, and the response time target of the consulting service. The cost optimization resource allocation suggestions include the optimization suggestions for drug procurement batches, inventory turnover rate targets, and the layout from the distribution center to the service point. The distribution path efficiency optimization suggestions include the priority distribution path for emergency drugs, the cost-effective route for regular distribution, and the distribution frequency adjustment. The drug demand trend analysis results include seasonal drug demand change charts, drug demand forecasts during disease outbreaks, and annual forecasts of drug use trends in the elderly population. The real-time dynamic adjustment strategies include critical alarm values ​​for drug inventory, key indicator thresholds for health status monitoring, and medication adjustment reminders based on changes in health data.

3. The method for managing medication for the elderly according to claim 1, characterized in that: Based on the decision support auxiliary information, the arrival and processing process of drug delivery requests is simulated, the waiting and service time of each request is recorded, and the human resource allocation is optimized according to the efficiency analysis. The specific steps of generating process optimization suggestions are as follows; Based on the decision support auxiliary information, the drug delivery process is mapped, an event-based simulation is performed, the arrival and processing of drug delivery requests are simulated, the waiting time and service time of each request are recorded, and the drug delivery process efficiency analysis results are generated; Based on the results of the drug distribution process efficiency analysis, conduct a human resource configuration analysis of the drug review process, simulate the review capabilities under various configurations, calculate the average review completion time and queue length, optimize the reviewer configuration, and generate drug review process configuration recommendations; Based on the drug review process configuration suggestions, combined with instant consultation request data, the number of consultation service windows is adjusted, the customer waiting time and service efficiency under various numbers of service windows are simulated, and process optimization suggestions are generated.

4. The method for managing medication for the elderly according to claim 1, characterized in that: Based on the cost-optimized resource allocation suggestion, a drug distribution network is constructed, cost and time weights are assigned to each distribution path, and a distribution route is selected according to cost efficiency. The specific steps of generating a distribution path efficiency optimization suggestion are as follows; Based on the cost-optimized resource allocation suggestion, a drug distribution network model is constructed, distribution centers and receiving points are defined as network nodes, distribution paths are defined as network edges, cost and time weights are assigned to each edge, and a drug distribution network construction result is generated; Based on the drug distribution network construction result, the lowest cost path from the distribution center to the receiving point is calculated, and the path selection and calculation are performed with reference to the carrying capacity and distribution time limit of the route to generate a distribution path selection result; Based on the distribution path selection results, the distribution efficiency is evaluated, the transportation process of the drugs from the distribution center to the receiving point is simulated, the transportation time and cost are recorded, the efficiency impact of the path selection is analyzed, and distribution path efficiency optimization suggestions are generated.

5. A drug management system for the elderly, characterized in that: According to any one of claims 1 to 4, the method for managing medication for the elderly is implemented, and the system comprises: The pharmacokinetic model analysis module is based on the medication data of the elderly. It analyzes the drug absorption, distribution, metabolism and excretion process through numerical simulation, sets the initial conditions and time steps, and generates auxiliary information for drug dynamic analysis. The decision optimization simulation module performs optimization simulation of drug management decisions based on the drug dynamic analysis auxiliary information, including drug selection, medication time and dosage adjustment, and generates decision support auxiliary information; The process optimization simulation module simulates the processes of drug distribution, review and consultation based on the decision support auxiliary information, optimizes waiting time and resource allocation, and generates process optimization suggestions; The cost optimization planning module designs a resource allocation plan for drug supply chain cost optimization based on the process optimization suggestions, including dynamic adjustment of inventory, dispensing and distribution priorities, and performs cost-benefit analysis to generate cost-optimized resource allocation suggestions; The path efficiency optimization module optimizes the drug distribution path and speed based on the cost-optimized resource allocation suggestion, evaluates the distribution efficiency, and generates a distribution path efficiency optimization suggestion; The demand prediction model module constructs a prediction model for future drug demand based on the drug dynamic analysis auxiliary information and the drug taking history data of the elderly, and performs demand trend analysis to generate drug demand trend analysis results; The real-time adjustment strategy module designs a dynamic adjustment strategy for drug inventory status and health indicators of elderly patients based on the drug demand trend analysis results and in combination with the real-time health data of the elderly, updates resource priorities, and generates a real-time dynamic adjustment strategy; The data integration module performs data integration operations based on the drug dynamic analysis auxiliary information, decision support auxiliary information, process optimization suggestions, cost-optimized resource allocation suggestions, distribution path efficiency optimization suggestions, drug demand trend analysis results and real-time dynamic adjustment strategies to generate a comprehensive data analysis set.

Images