Intelligent analysis system for public health decision and trend prediction

Abstract

The invention discloses an intelligent analysis system for public health decision and trend prediction. The intelligent analysis system comprises the following steps: establishing a random optimal control system state equation containing a drift term, a diffusion term and a control function, and an objective function; a disturbance term is introduced into the state equation, and an original optimal control problem is converted into a random differential game problem; constructing a smooth trajectory estimation function based on the observation data, and defining a joint objective function; solving a random differential game problem to obtain an optimal control function and a system parameter estimator; the method is applied to a random control SIRD infectious disease model, and dynamic estimation of time-varying parameters is achieved; and on the basis of an estimation result, predicting a future disease transmission trend, and providing decision support for prospective planning and accurate configuration of public health resources. According to the method, a system with high noise and time-varying parameters can be effectively processed, the prediction precision and robustness are improved, and a quantitative basis is provided for prevention and control of infectious diseases.

Description

Technical Field

[0001] This invention belongs to the field of public health modeling and prediction technology, and specifically relates to an intelligent identification and prediction system for high-dimensional, dynamic, and complex public health systems. More particularly, it relates to an integrated system that enables data uploading, visualization analysis, model training, and epidemic prediction via a web interface. This system can be widely applied in areas such as infectious disease transmission trend analysis, public health intervention effectiveness evaluation, and medical resource demand prediction, providing a scientific basis for public health decision-making. Background Technology

[0002] High-dimensional dynamic systems are widely found in fields such as engineering control, finance and economics, biology and ecology, and public health. These systems typically exhibit characteristics such as high state dimensionality, complex dynamic behavior, strong nonlinearity, and susceptibility to random noise. Achieving accurate intelligent identification and trend prediction of these systems has become a common challenge in the interdisciplinary research of systems science, control theory, and artificial intelligence. Traditional system identification and prediction methods have significant limitations in dealing with such problems: most are based on the assumption of constant parameters, failing to fully consider the time-varying behavior of parameters caused by the system's intrinsic characteristics or external environment; at the same time, traditional methods often treat parameter estimation and system control problems separately, ignoring the dynamic coupling relationship between the two. When faced with high-dimensional states, strong noise, and control inputs coexisting, such "step-by-step" or "decoupled" strategies are prone to problems such as accumulated estimation bias, weak model generalization ability, and suboptimal control performance, making it difficult to meet the high requirements of model accuracy and robustness in complex application scenarios.

[0003] Public health systems, as a typical high-dimensional dynamic system, encompass multiple dimensions such as population health status, disease transmission mechanisms, prevention and control interventions, and resource allocation. They are characterized by high nonlinearity, strong time-varying parameters, and susceptibility to random factors. Traditional modeling methods often treat system parameters as fixed, failing to effectively incorporate external dynamic factors such as policy adjustments and changes in social behavior. This leads to significant discrepancies between model predictions and actual evolution, thus hindering the improvement of accurate and efficient public health decision-making capabilities.

[0004] Taking infectious disease transmission systems as an example, diseases such as AIDS, SARS, COVID-19, monkeypox, Ebola virus, and tuberculosis continue to threaten public health and safety globally. Against this backdrop, traditional experience-based prevention and control strategies often lag behind the actual development of the epidemic, making it difficult to achieve optimal allocation of public health resources. Therefore, constructing dynamic models that accurately depict the transmission patterns of infectious diseases and rapidly and accurately estimating time-varying parameters such as transmission rate, infection rate, and recovery rate is not only theoretically valuable in revealing the mechanisms of epidemic evolution but also of significant practical importance in evaluating the effectiveness of prevention and control measures and predicting future trends. However, existing epidemic prediction and parameter estimation methods still face many challenges in handling highly nonlinear, noisy environments and time-varying parameters: traditional methods typically set parameters as constants, ignoring their evolution over time and failing to effectively integrate external dynamic information such as policy interventions and changes in social behavior. Furthermore, conventional parameter estimation processes are often separated from the control problem, failing to fully consider the synergistic mechanism between the two, resulting in insufficient estimation accuracy and robustness in real-world applications with high noise and strong uncertainty. Therefore, developing an intelligent identification and prediction method that can simultaneously achieve parameter estimation and system control, and has strong robustness to time-varying characteristics and random noise, has become a critical cross-disciplinary technical challenge that urgently needs to be overcome. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the existing technology and provide an intelligent analysis system for public health decision-making and trend prediction, comprising the following steps:

[0006] Step S1: Establish the state equation and objective function of the stochastic optimal control system. The state equation includes drift terms, diffusion terms, and a control function. The objective function includes operating cost and terminal cost.

[0007] Step S2: Introduce a disturbance term into the state equation to transform the original stochastic optimal control problem into a stochastic differential game problem;

[0008] Step S3: Construct a smooth trajectory estimation function based on the observation data, and define a joint objective function that includes trajectory fitting error and perturbation penalty;

[0009] Step S4: Solve the stochastic differential game problem using the Pontryagin maximum principle to obtain the optimal control function and system parameter estimates;

[0010] Step S5: Apply the estimation method to the stochastic control SIRD infectious disease model to achieve the estimation of time-varying parameters. Dynamic estimation;

[0011] Step S6: Predict the spread trend of infectious diseases based on estimated parameters.

[0012] Further, in step S1, the state equation and objective function of the stochastic optimal control system are established. The state equation includes a drift term, a diffusion term, and a control function. The objective function includes operating cost and terminal cost, as follows:

[0013]

[0014] Among them, stochastic processes exist middle, These are parameters to be estimated. It is the diffusion coefficient. It is Brownian motion. It is a control function. It is a performance metric function.

[0015] Furthermore, in step S2, a disturbance term is introduced into the state equation. The original stochastic optimal control problem is transformed into a stochastic differential game problem, and its system equations become:

[0016]

[0017] Among them, the function It is a linear perturbation.

[0018] Further, in step S3, a smooth trajectory estimation function is constructed based on the observed data, and a joint objective function including trajectory fitting error and perturbation penalty is defined. Specific steps include:

[0019] Step S31: Construct a smooth trajectory estimation function based on the observation data;

[0020]

[0021] Based on discrete observation data A continuous smooth trajectory estimate is constructed using spline interpolation. To eliminate observation noise and obtain trajectory estimates over continuous time;

[0022] Step S32: Define a joint objective function that includes trajectory fitting error and perturbation penalty;

[0023] For any smooth function and positive definite matrix We introduce the objective function as follows:

[0024]

[0025] Meanwhile, we define the joint objective function as:

[0026]

[0027] Finally, the estimator is defined as:

[0028]

[0029] Further, in step S4, the stochastic differential game problem is solved using the Pontryagin maximum principle to obtain the optimal control function and system parameter estimates. Specific steps include:

[0030] Step S41: Apply Pontryagin's maximum principle to solve the stochastic differential game problem, and construct and control the objective respectively. and disturbance targets The corresponding Hamiltonian function and ;

[0031] Step S42: By solving the necessary conditions derived from the maximum principle, including the state equations, co-state equations, and transversal conditions, the optimal control is obtained. and With system parameter estimates .

[0032] Furthermore, in step S5, the estimation method is applied to the stochastic SIRD infectious disease model to achieve the estimation of time-varying parameters. , and The dynamic estimation, specifically includes the following steps:

[0033] Step S51: Data preprocessing, specifically including:

[0034] Step S511: Obtain historical infectious disease data for the target area, including but not limited to: daily new infection numbers. Cumulative number of recovered patients Cumulative death toll and total population The data sources include statistical data published by public health departments;

[0035] Step S512: Clean and align the acquired data, handle missing and outlier values, and ensure the continuity and consistency of the data over time.

[0036] Step S513: Based on the cleaned data, calculate or initialize the susceptible population. This forms a complete time series dataset for parameter estimation. ;

[0037] Step S52: Based on the transmission mechanism of the target infectious disease, establish a stochastic differential equation model containing control variables as the infectious disease dynamics model. Specific steps include:

[0038] Step S521: The dynamic equations of the stochastic control SIRD model are established as follows:

[0039]

[0040] in, It is a control variable representing the vaccination rate. Indicates the rate of disease transmission. Indicates the cure rate. The mortality rate. It is standard Brownian motion. This indicates that the spread of the epidemic is influenced by many random factors. , and It is a parameter that changes over time;

[0041] Step S522: Define the control objective function to minimize the scale of infection, the scale of death, and the control cost;

[0042] Step S53: Based on the infectious disease dynamics model and actual observation data, construct a stochastic differential game problem, specifically including:

[0043] Step S531: Introduce a virtual second participant, adding a perturbation to the dynamic equations of the stochastically controlled SIRD model. :

[0044]

[0045] in , For inclusion The diffusion term;

[0046] Step S532: Define the objective function for the second participant.

[0047]

[0048] Step S533: At this point, the original parameter estimation problem has been transformed into a stochastic differential game problem, with the joint objective function being:

[0049]

[0050] Step S54: Solve the stochastic differential game problem using Pontryagin's maximum principle, deriving the boundary value problem composed of the state equation, co-state equation, and cross-sectional conditions, specifically including:

[0051] Step S541: Construct Hamiltonian functions for both sides of the game. and ;

[0052] Step S542: Apply Pontryagin's maximum principle, by letting and Thus, the analytical expression of the optimal control function is obtained;

[0053] Step S543: Derive the boundary value problem describing the optimal evolution trajectory of the system, which includes: state equations, co-state equations, and cross-cutting conditions;

[0054] Step S55: Numerically solve the boundary value problem to simultaneously obtain the optimal estimates of the time-varying parameters and the optimal control function in the infectious disease dynamics model.

[0055] Step S551: Use a numerical algorithm to solve the boundary value problem obtained in S544;

[0056] Step S552: During the numerical solution process, the time-varying parameters are... By substituting the parameterized form into the data and iteratively optimizing it, the obtained trajectory best fits the observed data.

[0057] Step S553: ​​The numerical solution process synchronously outputs the optimal estimated trajectory of the time-varying parameters and the optimal control strategy trajectory;

[0058] Step S56: Using the optimal estimate of the time-varying parameters and the optimal control function, predict the future spread trend of the target infectious disease:

[0059] Step S561: The estimated parameters Extrapolated values ​​within the forecast period and designed future control strategies Then substitute it into the stochastic control SIRD model established in step S521;

[0060] Step S562: Use numerical methods to perform multiple simulations of the future stochastic dynamic system to generate multiple possible propagation trajectories;

[0061] Step S563: Based on the simulation results, the mean prediction and confidence interval of the number of people in each cell are statistically obtained, thereby realizing the probabilistic prediction of the epidemic trend of infectious diseases and providing a quantitative basis for the allocation of medical resources.

[0062] Compared with the prior art, the significant advantages of the present invention are reflected in the following aspects:

[0063] 1. By introducing a disturbance term, the original stochastic optimal control problem is transformed into a stochastic differential game, thus achieving for the first time the simultaneous estimation of system parameters and the optimal control function.

[0064] 2. Unlike traditional methods that handle optimal control and parameter estimation separately, this invention fully considers the interaction between the two, enabling the model to dynamically adapt to changes in unknown parameters. This results in higher accuracy and robustness in complex nonlinear and high-noise environments.

[0065] 3. This invention fills a gap in the handling of time-varying parameter estimation in infectious disease models. It accurately captures the changing patterns of parameters over time and effectively integrates external influences such as multidimensional data, environmental factors, social behavior, and policy interventions, thereby overcoming the limitations of traditional models in propagation modeling within complex social contexts. This invention demonstrates excellent accuracy and robustness in time-varying parameter estimation, providing more reliable theoretical and methodological support for infectious disease prevention and control and policy formulation.

[0066] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. Attached Figure Description

[0067] Figure 1 This is a system flowchart of the present invention;

[0068] Figure 2 This is a schematic diagram of the main interface layout of a web system.

[0069] Figure 3 This is a schematic diagram of the trend analysis interface for a web system.

[0070] Figure 4 This is a schematic diagram of the Web system distribution analysis interface;

[0071] Figure 5 This is a schematic diagram of the Web system model training interface;

[0072] Figure 6 This is a schematic diagram of the epidemic prediction interface of the web system;

[0073] Figure 7 A schematic diagram illustrating the prediction of confidence intervals and prevention and control scenarios for web systems. Detailed Implementation

[0074] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Those skilled in the art can easily understand other advantages and effects of this disclosure from the content disclosed in this specification. Specifically, the embodiments of the present invention are implemented through a front-end interactive interface built on Web technology. This interface is developed using HTML5, CSS3, and JavaScript technologies and integrates the ECharts visualization library, providing users with a complete graphical operation process from data upload to model prediction. The accompanying drawings will show the various functional modules of the interface and their interaction processes. It should be emphasized that the described embodiments are only a part of the embodiments of this disclosure. The present invention can also be applied through other different implementation methods. Without departing from the basic concept of the present invention, the details in this specification can also be adjusted or changed according to different viewpoints and applications. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0075] As shown in the figure, an embodiment of the present invention provides an intelligent analysis system for public health decision-making and trend prediction, comprising the following steps:

[0076] Step S1: Establish the state equation and objective function of the stochastic optimal control system. The state equation includes drift terms, diffusion terms, and a control function. The objective function includes operating cost and terminal cost.

[0077] Step S2: Introduce a disturbance term into the state equation to transform the original stochastic optimal control problem into a stochastic differential game problem;

[0078] Step S3: Construct a smooth trajectory estimation function based on the observation data, and define a joint objective function that includes trajectory fitting error and perturbation penalty;

[0079] Step S4: Solve the stochastic differential game problem using the Pontryagin maximum principle to obtain the optimal control function and system parameter estimates;

[0080] Step S5: Apply the estimation method to the stochastic SIRD infectious disease model to achieve dynamic estimation of time-varying parameters β(t), γ(t), and μ(t);

[0081] Step S6: Predict the spread trend of infectious diseases based on estimated parameters.

[0082] Further, in step S1, the state equation and objective function of the stochastic optimal control system are established. The state equation includes a drift term, a diffusion term, and a control function. The objective function includes operating cost and terminal cost, as follows:

[0083]

[0084] Among them, stochastic processes exist middle, These are parameters to be estimated. It is the diffusion coefficient. It is Brownian motion. It is a control function. It is a performance metric function.

[0085] Furthermore, in step S2, a disturbance term is introduced into the state equation. The original stochastic optimal control problem is transformed into a stochastic differential game problem, and its system equations become:

[0086]

[0087] Among them, the function It is a linear perturbation.

[0088] Further, in step S3, a smooth trajectory estimation function is constructed based on the observed data, and a joint objective function including trajectory fitting error and perturbation penalty is defined. Specific steps include:

[0089] Step S31: Construct a smooth trajectory estimation function based on the observation data;

[0090]

[0091] Based on discrete observation data A continuous smooth trajectory estimate is constructed using spline interpolation. To eliminate observation noise and obtain trajectory estimates over continuous time;

[0092] Step S32: Define a joint objective function that includes trajectory fitting error and perturbation penalty;

[0093] For any smooth function and positive definite matrix We introduce the objective function as follows:

[0094]

[0095] Meanwhile, we define the joint objective function as:

[0096]

[0097] Finally, the estimator is defined as:

[0098] .

[0099] Further, in step S4, the stochastic differential game problem is solved using the Pontryagin maximum principle to obtain the optimal control function and system parameter estimates. Specific steps include:

[0100] Step S41: Apply Pontryagin's maximum principle to solve the stochastic differential game problem, and construct and control the objective respectively. and disturbance targets The corresponding Hamiltonian function and ;

[0101] Step S42: By solving the necessary conditions derived from the maximum principle, including the state equations, co-state equations, and transversal conditions, the optimal control is obtained. and With system parameter estimates .

[0102] Furthermore, in step S5, the estimation method is applied to the stochastic SIRD infectious disease model to achieve the estimation of time-varying parameters. , and The dynamic estimation, specifically includes the following steps:

[0103] Step S51: Data preprocessing, specifically including:

[0104] Step S511: Obtain historical infectious disease data for the target area, including but not limited to: daily new infection numbers. Cumulative number of recovered patients Cumulative death toll and total population The data sources include statistical data published by public health departments;

[0105] Step S512: Clean and align the acquired data, handle missing and outlier values, and ensure the continuity and consistency of the data over time.

[0106] Step S513: Based on the cleaned data, calculate or initialize the susceptible population. This forms a complete time series dataset for parameter estimation. ;

[0107] Step S52: Based on the transmission mechanism of the target infectious disease, establish a stochastic differential equation model containing control variables as the infectious disease dynamics model. Specific steps include:

[0108] Step S521: The dynamic equations of the stochastic control SIRD model are established as follows:

[0109]

[0110] in, It is a control variable representing the vaccination rate. Indicates the rate of disease transmission. Indicates the cure rate. The mortality rate. It is standard Brownian motion. This indicates that the spread of the epidemic is influenced by many random factors. , and It is a parameter that changes over time;

[0111] Step S522: Define the control objective function to minimize the scale of infection, the scale of death, and the control cost;

[0112] Step S53: Based on the infectious disease dynamics model and actual observation data, construct a stochastic differential game problem, specifically including:

[0113] Step S531: Introduce a virtual second participant, adding a perturbation to the dynamic equations of the stochastically controlled SIRD model. :

[0114]

[0115] in , For inclusion The diffusion term;

[0116] Step S532: Define the objective function for the second participant.

[0117]

[0118] Step S533: At this point, the original parameter estimation problem has been transformed into a stochastic differential game problem, with the joint objective function being:

[0119] ;

[0120] Step S54: Solve the stochastic differential game problem using Pontryagin's maximum principle, deriving the boundary value problem composed of the state equation, co-state equation, and cross-sectional conditions, specifically including:

[0121] Step S541: Construct Hamiltonian functions for both sides of the game. and ;

[0122] Step S542: Apply Pontryagin's maximum principle, by letting and Thus, the analytical expression of the optimal control function is obtained;

[0123] Step S543: Derive the boundary value problem describing the optimal evolution trajectory of the system, which includes: state equations, co-state equations, and cross-cutting conditions;

[0124] Step S55: Numerically solve the boundary value problem to simultaneously obtain the optimal estimates of the time-varying parameters and the optimal control function in the infectious disease dynamics model.

[0125] Step S551: Use a numerical algorithm to solve the boundary value problem obtained in S544;

[0126] Step S552: During the numerical solution process, the time-varying parameters are... By substituting the parameterized form into the data and iteratively optimizing it, the obtained trajectory best fits the observed data.

[0127] Step S553: ​​The numerical solution process synchronously outputs the optimal estimated trajectory of the time-varying parameters and the optimal control strategy trajectory;

[0128] Step S56: Using the optimal estimate of the time-varying parameters and the optimal control function, predict the future spread trend of the target infectious disease:

[0129] Step S561: The estimated parameters Extrapolated values ​​within the forecast period and designed future control strategies Then substitute it into the stochastic control SIRD model established in step S521;

[0130] Step S562: Use numerical methods to perform multiple simulations of the future stochastic dynamic system to generate multiple possible propagation trajectories;

[0131] Step S563: Based on the simulation results, the mean prediction and confidence interval of the number of people in each cell are statistically obtained, thereby realizing the probabilistic prediction of the epidemic trend of infectious diseases and providing a quantitative basis for the allocation of medical resources.

[0132] The above embodiments are provided merely for the purpose of describing the present invention and are not intended to limit the scope of the invention. The scope of the invention is defined by the appended claims. Various equivalent substitutions and modifications made without departing from the spirit and principles of the invention should be covered within the scope of the invention.

Claims

1. characterized in that, The method comprises the following steps: Step S1: establishing a state equation and an objective function of a random optimal control system, the state equation comprising a drift term, a diffusion term and a control function, and the objective function comprising a running cost and a terminal cost; Step S2: introducing a disturbance term in the state equation to convert the original random optimal control problem into a random differential game problem; Step S3: constructing a smooth trajectory estimation function based on observation data, and defining a joint objective function comprising a trajectory fitting error and a disturbance penalty; Step S4: solving the random differential game problem to obtain an optimal control function and a system parameter estimate; Step S5: Applying the estimation method to the stochastic control SIRD epidemic model to achieve dynamic estimation of the time-varying parameters ; Step S6: predicting a future disease transmission trend based on the estimation result, and providing decision support for prospective planning and precise allocation of public health resources.

2. The intelligent analysis system for public health decision and trend prediction according to claim 1, characterized in that, The step S1: establishing a state equation and an objective function of a random optimal control system, the state equation comprising a drift term, a diffusion term and a control function, and the objective function comprising a running cost and a terminal cost, specifically comprises: ; where the random process In which, is the parameter to be estimated, is the diffusion coefficient, is the Brownian motion; is the control function; is the performance index function. 3.The intelligent analysis system for public health decision and trend prediction of claim 1, wherein, The step S2: introducing a perturbation term in the state equation The original stochastic optimal control problem is converted into a stochastic differential game problem, and the system equation becomes: ; where the function is a linear perturbation.

4. The intelligent analysis system for public health decision and trend prediction according to claim 1, characterized in that, The step S3: constructing a smooth trajectory estimation function based on observation data, and defining a joint objective function comprising a trajectory fitting error and a disturbance penalty, specifically comprises: Step S31: constructing a smooth trajectory estimation function based on observation data, comprising: ； Based on discrete observation data A continuous smooth trajectory estimate is constructed by a spline interpolation method to eliminate observation noise and obtain a trajectory estimate on continuous time; Step S32: defining a joint objective function comprising a trajectory fitting error and a disturbance penalty, comprising: For any smooth function and positive definite matrix we introduce the objective function ; At the same time, the joint objective function is defined as ; Finally, the estimate is defined as 。 5. The intelligent analysis system for public health decision and trend prediction according to claim 1, characterized in that, The step S4: solving the random differential game problem to obtain an optimal control function and a system parameter estimate, specifically comprises: Step S41: solving the stochastic differential game problem, respectively constructing control target and disturbance target corresponding Hamilton function and ; Step S42: Obtain the optimal control by solving the necessary conditions derived from the maximum principle, including the state equation, the co-state equation and the transversality condition and and the system parameter estimator .

6. The intelligent analysis system for public health decision and trend prediction according to claim 1, characterized in that, The step S5: applying an estimation method to the random SIRD epidemic model, realizing dynamic estimation of the time-varying parameters , and , specifically comprising: Step S51: data preprocessing, specifically comprising: Step S511: Obtain historical infectious disease data of the target region, including but not limited to: daily new infection number , cumulative recovery number , cumulative death number , and total population number ; the data sources include statistical data published by public health departments; Step S512: cleaning and aligning the obtained data, processing missing values and outliers, and ensuring the continuity and consistency of the data in the time series; Step S513: Based on the cleaned data, calculate or initialize the susceptible population , forming a complete time series dataset for parameter estimation ; Step S52: establishing a random differential equation model comprising a control variable as an infectious disease dynamics model according to the transmission mechanism of the target infectious disease, specifically comprising: Step S521: establishing a dynamics equation of a random control SIRD model, ; wherein, is a control variable representing vaccination rate; represents the spread rate of the disease, represents the cure rate, is the mortality rate; is a standard Brownian motion, represents that the spread of the epidemic is affected by many random factors. , and are time-varying parameters; Step S522: defining a control objective function to minimize the infection size, the death size and the control cost: ; Step S53: based on the infectious disease dynamics model and the actual observation data, constructing a random differential game problem, specifically comprising: Step S531: Introducing a virtual second participant, adding a perturbation in the dynamics equation of the stochastic control SIRD model : ; wherein , is a diffusion term comprising ; Step S532: defining the objective function of the second participant, ; Step S533: at this point, the original parameter estimation problem is converted into a random differential game problem, and the joint objective function is: ； Step S54: solving the random differential game problem, and deriving a boundary value problem composed of a state equation, a co-state equation and a cross-section condition, specifically comprising: Step S541: Construct Hamiltonian function for both sides of the game and ; Step S542: solving the first order optimality conditions of the stochastic differential game model by letting and , to obtain the analytical expression of the optimal control function; Step S543: deriving a boundary value problem describing the optimal evolution trajectory of the system, which comprises: (1) a state equation; (2) a co-state equation, which is a differential equation evolving in reverse time; (3) a cross-section condition, which is a value condition of the co-state at the terminal time; Step S55: numerically solving the boundary value problem to simultaneously obtain an optimal estimate value of a time-varying parameter in the infectious disease dynamics model and an optimal control function: Step S551: using a numerical algorithm to solve the boundary value problem obtained in S544; Step S552: During the numerical solution process, the time-varying parameters are substituted into the parameterized form, and through iterative optimization, the trajectory obtained by the solution is best fitted to the observation data. Step S553: The process of numerical solution synchronously outputs the optimal estimation trajectory of the time-varying parameter and the optimal control strategy trajectory; Step S56: The future transmission trend of the target infectious disease is predicted by using the optimal estimation value of the time-varying parameter and the optimal control function: Step S561: Substitute the estimated parameters into the SIRD model established in step S521 extrapolated values in the prediction period and the designed future control strategy , into the SIRD model established in step S521. Step S562: A plurality of simulations of the future random dynamic system are performed by using a numerical method to generate a plurality of possible transmission trajectories; Step S563: Based on the simulation results, the mean prediction and the confidence interval of the population number of each compartment in the future are statistically obtained, so as to realize the probabilistic prediction of the epidemic trend of the infectious disease and provide a quantitative basis for the allocation of medical resources.

Images