Medical resource optimization analysis system based on Boston matrix model

Abstract

The present invention discloses a medical resource optimization and analysis system based on the Boston matrix model, and relates to the technical field of data analysis; the system comprises: a data acquisition and preprocessing unit, a Boston matrix construction and analysis unit, and a genetic algorithm optimization unit; the data acquisition and processing unit is used to collect medical data from various data sources and generate a high-dimensional feature matrix; the Boston matrix construction and analysis unit is used to construct a Boston matrix according to the high-dimensional feature matrix, classify different medical resources, and obtain classification results; the genetic algorithm optimization unit uses a genetic algorithm to optimize the allocation of medical resources according to the classification results, and when the iterative optimization reaches a set number of times, the final resource allocation vector is output as the optimal configuration of medical resources. The present invention can improve the efficiency and effect of resource allocation, and can also maintain high flexibility and adaptability in a complex and changeable medical environment.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and in particular to a medical resource optimization analysis system based on a Boston matrix model. Background Art

[0002] Medical resources usually include hospital beds, medical equipment, medical staff, medicines, etc. The optimal allocation of these resources is of great significance to improving the quality of medical services, reducing medical costs and improving patient satisfaction. In the existing technology, there are various methods for the allocation and management of medical resources. However, these methods still have many problems to be solved in practical applications and cannot fully meet the needs of the modern medical system.

[0003] In the prior art, the allocation of medical resources is mostly based on traditional statistical analysis methods, such as linear programming, integer programming, and heuristic algorithms. These methods mainly rely on linear analysis and static modeling of historical data, and calculate the optimal resource allocation plan by formulating resource allocation rules and optimization goals. For example, the linear programming method constructs an objective function and constraints to find a resource allocation plan that maximizes or minimizes the objective function. This method has certain advantages in a static, deterministic resource allocation environment and can provide a reasonable resource allocation plan. However, in an actual medical environment, resource requirements are often dynamic and complex, and linear programming and other traditional methods are difficult to adapt to such dynamic changes. In addition, when these methods deal with complex resource allocation problems with multiple dimensions and multiple constraints, the computational complexity is high and the solution efficiency is low, and it is often difficult to meet the real-time requirements in practical applications.

[0004] To solve the above problems, in recent years, some studies have introduced machine learning and data-driven methods to optimize the allocation of medical resources. For example, some studies use methods based on regression analysis and prediction models to predict future medical resource needs and dynamically allocate resources based on the prediction results. This method establishes a data-driven prediction model by analyzing historical data, which improves the flexibility and adaptability of resource allocation to a certain extent. However, these methods mainly focus on the accuracy of resource demand prediction, and lack effective solutions for how to make optimal allocation under complex constraints. In addition, data-driven prediction models often have high requirements for data quality and sample size. Once the data quality is poor or the sample size is insufficient, the reliability and accuracy of the prediction results will be greatly reduced, which will affect the final effect of resource allocation. In addition, some existing technologies try to use intelligent optimization algorithms, such as genetic algorithms, particle swarm optimization algorithms, and simulated annealing algorithms, to solve the problem of medical resource allocation. Intelligent optimization algorithms simulate the evolutionary process of nature or the annealing process of physical systems to find the global optimal solution to complex optimization problems. In medical resource optimization, these algorithms can make up for the shortcomings of traditional methods to a certain extent, especially in multi-objective and multi-constraint optimization problems. However, these intelligent optimization algorithms also face some challenges in practical applications. First, the computational complexity of the algorithm is high, especially for large-scale, multi-dimensional problems, the solution time is often unacceptable. Second, these algorithms have a great influence on the final optimization results in terms of initial population selection and parameter setting. If the initial population is not properly selected or the parameters are not set reasonably, the algorithm may fall into a local optimal solution. In addition, these algorithms usually rely on certain heuristic strategies, which often lack theoretical basis and versatility, and perform unevenly in different application scenarios. Summary of the invention

[0005] The purpose of the present invention is to provide a medical resource optimization analysis system based on the Boston Matrix model, which can not only improve the efficiency and effectiveness of resource allocation, but also maintain high flexibility and adaptability in a complex and changeable medical environment, providing a scientific and effective technical means for improving the quality of medical services and optimizing resource management.

[0006] In order to solve the above technical problems, the present invention provides a medical resource optimization and analysis system based on the Boston matrix model, the system comprising: a data acquisition and preprocessing unit, a Boston matrix construction and analysis unit and a genetic algorithm optimization unit; the data acquisition and processing unit is used to collect medical data from various data sources, delete redundant and erroneous data in the medical data, fill in missing values, extract features after nonlinear transformation of the medical data, and then use a nonlinear mapping function to perform high-dimensional mapping on the features to generate a high-dimensional feature matrix; the Boston matrix construction and analysis unit is used to calculate the market growth rate and market share of each medical resource according to the high-dimensional feature matrix, construct a Boston matrix, classify different medical resources, and obtain classification results; the genetic algorithm optimization unit uses a genetic algorithm to optimize the allocation of medical resources according to the classification results, and continuously iterates and optimizes the allocation plan of medical resources through selection, crossover and mutation operations. When the iterative optimization reaches a set number of times, the final resource allocation vector is output as the optimal configuration of medical resources.

[0007] Furthermore, the data acquisition and preprocessing unit includes: a data acquisition module and a preprocessing module; the data acquisition module collects multi-dimensional medical data to form a data set:

[0008] ,

[0009] Each data point Corresponding medical resources data points, is the total number of medical resources, including the following attributes: Number of patients , which represents the number of patients treated in a given period of time and the utilization rate of medical equipment , indicating the frequency of use and disease distribution of medical equipment in a given period of time , which represents the proportion of patients with different diseases and other related variables in a given period of time ; Including: bed occupancy rate, drug usage, doctor working hours and / or nurse working hours; the preprocessing unit is used to delete redundant and erroneous data in the data set, fill in missing values, extract features after nonlinear transformation of the data set, and then use nonlinear mapping function to perform high-dimensional mapping on the features to generate a high-dimensional feature matrix .

[0010] Furthermore, the Boston matrix construction and analysis unit calculates the market growth rate of each medical resource according to the high-dimensional feature matrix, which specifically includes: Perform eigenvalue decomposition:

[0011] ;

[0012] in, is the eigenvector matrix; is a diagonal eigenvalue matrix; using the eigenvalues ​​and eigenvectors in the eigenvalue decomposition results, combined with the dynamic characteristics of resources, the growth rate calculation formula is constructed:

[0013] ;

[0014] in, For resources market growth rate; is the total number of resources; For the The characteristic value represents the contribution of resources in a specific characteristic direction; is the eigenvector matrix Middle Line Elements of the column, representing resources In the The projection in the feature direction; For resources The time derivative of the disease distribution indicates the speed of change of disease distribution and reflects the dynamic characteristics of market growth; represents a nonlinear transformation function; It is an adjustment parameter used to control the influence of the characteristic value on the growth rate and is a set value.

[0015] Furthermore, the Boston matrix construction and analysis unit calculates the market share of each medical resource according to the high-dimensional feature matrix, and the process specifically includes:

[0016] ;

[0017] in, For resources market share; is the eigenvector matrix Middle Line Elements of a column; For the eigenvalues; is the standard deviation of the distribution, which is used to adjust the impact of eigenvectors and eigenvalues ​​on market share; is the total number of patients.

[0018] Furthermore, the Boston matrix construction and analysis unit constructs the Boston matrix through the following formula to classify different medical resources and obtain classification results:

[0019] ;

[0020] in, For resources Position score in Boston matrix; is the base of natural logarithms; medical resources The classification results obtained for:

[0021] ;

[0022] in, and They are the upper and lower limits of the first classification threshold, respectively, used to distinguish the Boston Matrix score boundaries of different quadrants; and are the upper and lower limits of the second classification threshold, which are used to distinguish the market growth rate boundaries of different quadrants; , , and are four different preset positive integer values, representing the advantage diseases, cash diseases, problem diseases and disadvantage diseases in the Boston matrix model, satisfying: ,and .

[0023] Furthermore, the genetic algorithm optimization unit, when optimizing the allocation of medical resources using the genetic algorithm according to the classification result, initializes the population, each individual in the population represents a resource allocation scheme, and each scheme is represented by a vector Indicates that It is a resource In the The amount distributed among individuals;

[0024] ;

[0025] in: and are the upper and lower bounds of resource allocation respectively; To generate arrive A random number between .

[0026] Furthermore, when the genetic algorithm optimization unit uses the genetic algorithm to optimize the allocation of medical resources according to the classification results, the fitness function formula is as follows:

[0027] ;

[0028] in: For the The fitness value of each individual; For resources Boston Matrix score; For the Individual resources The amount of distribution; is the penalty coefficient, used to control over-allocation of resources; It is a regulation parameter used to adjust the impact of overall resource allocation deviation on fitness; Variance allocated for overall resources; is the total resource volume.

[0029] Furthermore, when the genetic algorithm optimization unit uses the genetic algorithm to optimize the allocation of medical resources according to the classification results, the selection probability formula of the selection operation is as follows:

[0030] ;

[0031] in, For the The probability of an individual being selected; For the The fitness of each individual; is the size of the population; the formula for the crossover operation is as follows:

[0032] ;

[0033] in, Resources for new individuals The amount of distribution; and Parent individuals and Chinese Resources The amount of distribution; is the crossover ratio, with a value of arrive between; is a random angle value used to introduce nonlinear changes; the formula for the mutation operation is as follows:

[0034] ;

[0035] in, The resources of the new individuals after mutation The amount of distribution; The resources of the new individuals generated after crossover The amount of distribution; is the coefficient of variation, controlling the amplitude of variation; is a standard normally distributed random variable; To adjust the parameters, the control variation range changes with the deviation of market share.

[0036] Furthermore, the genetic algorithm optimization unit uses the Lagrange multiplier method to constrain when continuously iterating and optimizing the allocation plan of medical resources. The formula is as follows:

[0037] ;

[0038] ;

[0039] in, is the Lagrange multiplier; Assigning vectors to resources; is the constraint function, which is expressed using the following formula:

[0040] .

[0041] The medical resource optimization analysis system based on the Boston matrix model of the present invention has the following beneficial effects:

[0042] The present invention innovatively applies the Boston Matrix model to the optimization of medical resources, and strategically classifies medical resources through two key indicators: market growth rate and market share. Traditional medical resource allocation usually relies on a single utilization rate or demand forecast, which ignores the comprehensive performance and future potential of resources in the market. The Boston Matrix model classifies and evaluates different types of resources through four quadrants: "advantage diseases", "cash diseases", "problem diseases" and "disadvantage diseases". For star resources (high growth rate, high market share), the system can recommend increasing investment to further expand the advantage; for cash disease resources (low growth rate, high market share), it can recommend maintaining or optimizing their current configuration; for problem resources (high growth rate, low market share) and disadvantaged disease resources (low growth rate, low market share), the system can propose a reduction or reconfiguration strategy. In this way, the system can focus on its long-term strategic significance while considering the current value of resources, and achieve the optimal strategy for resource allocation.

[0043] Secondly, the present invention combines the genetic algorithm in the intelligent optimization algorithm, which significantly improves the optimization effect and computational efficiency of resource allocation. Genetic algorithm is an optimization algorithm based on the principle of natural selection and genetic variation, and is particularly suitable for solving complex optimization problems with multiple objectives and multiple constraints. By encoding the resource allocation scheme, the genetic algorithm can quickly search for an approximate optimal solution in a huge solution space. The present invention uses the fitness function of the genetic algorithm to comprehensively evaluate the market performance of resources (such as market growth rate and market share) and the allocation effect. The fitness function is reasonably designed and can dynamically adjust weights and preferences to adapt to different medical needs and resource constraints. Through selection, crossover and mutation operations, the genetic algorithm continuously optimizes the resource allocation scheme in each generation of iteration and gradually approaches the optimal solution. Compared with traditional optimization methods, the genetic algorithm not only has higher computational efficiency, but also can avoid falling into local optimality, and has strong robustness and adaptability.

[0044] In addition, the present invention effectively solves the complex constraint problem in resource allocation by introducing the Lagrange multiplier method. In the actual allocation of medical resources, the total amount of resources, the maximum and minimum available amount of a single resource, the priority allocation strategy in emergency situations, etc. are all constraints that must be considered. Traditional optimization algorithms often face the problems of high computational complexity and difficulty in solving when dealing with these diverse and complex constraints. The Lagrange multiplier method makes the optimization process more concise and efficient by integrating the constraints into the objective function. By solving the partial derivatives of the Lagrange function, the present invention can optimize the resource allocation strategy and ensure the rationality and operability of resource allocation under the premise of ensuring that all constraints are met. This method not only improves the optimization capability of the system, but also enhances its flexibility and adaptability in practical applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0046] Figure 1 A schematic diagram of the system structure of a medical resource optimization and analysis system based on the Boston Matrix model provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0047] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0048] Example 1: Reference Figure 1 , a medical resource optimization and analysis system based on the Boston matrix model, the system includes: a data acquisition and preprocessing unit, a Boston matrix construction and analysis unit and a genetic algorithm optimization unit; the data acquisition and processing unit is used to collect medical data from various data sources, delete redundant and erroneous data in the medical data, fill in missing values, extract features after nonlinear transformation of the medical data, and then use a nonlinear mapping function to perform high-dimensional mapping on the features to generate a high-dimensional feature matrix; the Boston matrix construction and analysis unit is used to calculate the market growth rate and market share of each medical resource according to the high-dimensional feature matrix, construct a Boston matrix, classify different medical resources, and obtain classification results; the genetic algorithm optimization unit uses a genetic algorithm to optimize the allocation of medical resources according to the classification results, and continuously iterates and optimizes the allocation plan of medical resources through selection, crossover and mutation operations. When the iterative optimization reaches a set number of times, the final resource allocation vector is output as the optimal configuration of medical resources.

[0049] Specifically, the principle of the medical resource optimization and analysis system based on the Boston matrix model in Example 1 is to achieve refined management and dynamic optimization configuration of medical resources by integrating a variety of advanced data analysis and optimization technologies. The design goal of this system is to solve the problems of inefficiency, waste of resources, and inability to respond quickly to changing needs that are prevalent in the current allocation of medical resources. In the field of medical services, the effective allocation of resources is crucial. It not only affects the operational efficiency of medical institutions, but is also directly related to the patient's medical experience and health outcomes. Therefore, the medical resource optimization and analysis system based on the Boston matrix model systematically analyzes and optimizes the use of medical resources in a data-driven manner, aiming to achieve the optimal allocation of resources, thereby improving the quality and efficiency of medical services. The core of the system consists of four functional modules: a data acquisition and preprocessing unit, a Boston matrix construction and analysis unit, and a transmission algorithm optimization unit. First, the data acquisition and preprocessing unit is the basic module of the entire system, and its main function is to automatically collect medical-related data from various data sources. These data sources include but are not limited to hospital information systems (HIS), electronic medical records (EMR) systems, medical equipment monitoring data, government health statistics, etc. The types of data collected include patient diagnosis and treatment information, disease spectrum distribution, frequency of medical resource use, bed occupancy rate, equipment utilization rate, medical staff workload, etc. Through these rich multi-source data, the system can obtain a comprehensive and in-depth picture of the current status of medical services.

[0050] After data collection is completed, the system cleans and preprocesses the raw data. The purpose of data cleaning is to remove redundant, erroneous and incomplete data to improve data quality. Data preprocessing includes data standardization, normalization and denoising to ensure the accuracy and stability of subsequent analysis. For example, for missing data, the system can use interpolation methods to fill in; for outliers, the system sets thresholds or uses more complex statistical methods to identify and process them. More importantly, the system uses nonlinear transformations (such as logarithmic transformations or Box-Cox transformations) to reduce the skewed distribution of data and improve the normality of data, so as to better extract feature information. The extracted feature information undergoes further nonlinear mapping, such as high-dimensional mapping using Gaussian kernel functions or deep neural networks to generate a high-dimensional feature matrix. This high-dimensional feature matrix can better capture the complex relationships and patterns in the data, laying a data foundation for subsequent Boston matrix analysis. The Boston matrix construction and analysis unit is the second core module of the system. Its task is to calculate the market growth rate and market share of each medical resource based on the high-dimensional feature matrix generated above, and to construct the Boston matrix based on these indicators. The Boston matrix is ​​a classic strategic analysis tool, commonly used for product portfolio management of enterprises. In this system, the Boston Matrix is ​​cleverly applied to the classification and management of medical resources. Specifically, the system calculates the market growth rate and market share of each medical resource (such as hospital beds, medical staff, medical equipment, etc.). The market growth rate reflects the growth rate of resource demand, and the market share represents the proportion of resources in the entire medical resource allocation. Based on these two key indicators, the system divides medical resources into four categories: star type (high market growth rate, high market share), cash disease type (low market growth rate, high market share), problem type (high market growth rate, low market share) and disadvantaged disease type (low market growth rate, low market share). This classification method helps managers identify and decide which resources should be invested in and optimized, and which resources should be phased out or reallocated.

[0051] After the construction of the Boston matrix is ​​completed, the system enters the operation stage of the genetic algorithm optimization unit. Genetic algorithm is an optimization algorithm based on natural selection and genetic mechanism, which is particularly suitable for solving complex combinatorial optimization problems. In this system, the introduction of genetic algorithm is to further optimize the allocation plan of medical resources based on the Boston matrix analysis. The genetic algorithm optimization process usually starts with an initial population, and each individual represents a possible resource allocation plan. The system evaluates the pros and cons of each individual by designing a fitness function, which usually combines multidimensional objectives such as resource utilization efficiency, cost-effectiveness, and patient satisfaction. Then, the system retains individuals with higher fitness through selection operations, generates new resource allocation plans through crossover operations, and increases the diversity of the population through mutation operations to prevent the optimization from falling into local optimality. This process is iterated continuously until the set termination condition (such as the maximum number of iterations or the convergence of the fitness function) is reached, and finally the optimal resource allocation vector is output as the optimal configuration plan for medical resources. Finally, according to the optimization results output by the genetic algorithm, the allocation strategy of medical resources is adjusted in real time.

[0052] Embodiment 2: The data acquisition and preprocessing unit comprises: a data acquisition module and a preprocessing module; the data acquisition module collects multi-dimensional medical data to form a data set to form a data set:

[0053] ,

[0054] Each data point Corresponding medical resources data points, is the total number of medical resources, including the following attributes: Number of patients , which represents the number of patients treated in a given period of time and the utilization rate of medical equipment , indicating the frequency of use and disease distribution of medical equipment in a given period of time , which represents the proportion of patients with different diseases and other related variables in a given period of time ; Including: bed occupancy rate, drug usage, doctor working hours and / or nurse working hours; the preprocessing unit is used to delete redundant and erroneous data in the data set, fill in missing values, extract features after nonlinear transformation of the data set, and then use nonlinear mapping function to perform high-dimensional mapping on the features to generate a high-dimensional feature matrix .

[0055] Specifically, the data acquisition module undertakes the task of extracting key information from multi-dimensional and multi-source data. In modern medical systems, the sources of data are extremely diverse, including patient information systems, medical equipment monitoring, doctors' and nurses' work logs, and drug use records. These data not only reflect the use of medical resources, but also reveal the patient's treatment process, the load status of equipment, and the overall operation of medical services. Through the systematic collection and organization of these data, the system can form a comprehensive, multi-dimensional data set, laying a solid foundation for the optimal allocation of resources. However, the original data often contains redundant, inaccurate or incomplete information, which requires the intervention of the preprocessing module. The preprocessing module removes redundant and erroneous data in the data set through a series of data cleaning operations to ensure the accuracy and consistency of the data. At the same time, the preprocessing module uses advanced missing value filling technology to avoid analytical deviations caused by missing data. These operations not only improve the quality of the data, but also provide more accurate input for subsequent feature extraction and analysis. Next, the preprocessing module performs nonlinear transformation on the cleaned data. The purpose of nonlinear transformation is to better reveal the potential relationships and patterns in the data, especially in medical data, where there may be complex nonlinear relationships between many variables, which are difficult to capture through simple linear analysis. Through appropriate nonlinear transformations, the system can transform these complex relationships into a form that is easier to analyze and extract representative features. After extracting preliminary features, the system further performs high-dimensional mapping on these features. The core principle of high-dimensional mapping is to map data points that may be difficult to distinguish in low-dimensional space to high-dimensional space, making the relationship between these data points clearer and separable. This process usually uses nonlinear mapping functions, such as kernel methods or deep learning algorithms, to generate a high-dimensional feature matrix. The generation of a high-dimensional feature matrix is ​​the key to the system's efficient classification and optimization. It contains complex and important information about medical resources in multiple dimensions, and can provide accurate data support for the construction of the Boston Matrix model.

[0056] Embodiment 3: The process of the Boston matrix construction and analysis unit calculating the market growth rate of each medical resource according to the high-dimensional feature matrix specifically includes: Perform eigenvalue decomposition:

[0057] ;

[0058] in, is the eigenvector matrix; is a diagonal eigenvalue matrix; using the eigenvalues ​​and eigenvectors in the eigenvalue decomposition results, combined with the dynamic characteristics of resources, the growth rate calculation formula is constructed:

[0059] ;

[0060] in, For resources market growth rate; is the total number of resources; For the The characteristic value represents the contribution of resources in a specific characteristic direction; is the eigenvector matrix Middle Line Elements of the column, representing resources In the The projection in the feature direction; For resources The time derivative of the disease distribution indicates the speed of change of disease distribution and reflects the dynamic characteristics of market growth; represents a nonlinear transformation function; It is an adjustment parameter used to control the influence of the characteristic value on the growth rate and is a set value.

[0061] Specifically, the derivation of the formula first relies on the high-dimensional feature matrix The eigenvalue decomposition of Here, the matrix It represents the feature matrix formed after data preprocessing and high-dimensional mapping, which contains the performance of all medical resources in different feature dimensions. Eigenvalue decomposition decomposes this complex matrix into eigenvector matrix and the diagonal eigenvalue matrix ,in The column vector of the matrix The characteristic vector of The diagonal elements of are the corresponding eigenvalues. The eigenvectors provide the characteristic directions in the high-dimensional space, while the eigenvalues ​​indicate the importance or contribution of these characteristic directions. Through this decomposition method, the system can identify the characteristic directions with significant influence in the data and quantify their impact on the overall data. In the actual market growth rate calculation, the formula determines the market growth potential of each medical resource by integrating these eigenvalues ​​and eigenvectors. Specifically, the resource Market growth rate The calculation formula combines the eigenvalues and the elements of the eigenvector In the formula, the eigenvalue It represents the contribution of resources in a specific characteristic direction, and its size reflects the influence of this direction on the resource market performance. Elements in Representation Resources In the feature direction The projection on the direction reflects the performance strength of the resource in that direction. By taking the weighted sum of the projections of each resource in all feature directions, the formula can capture the performance of the resource in the overall market. In particular, using the sine function Processing the values ​​of feature vector projections, this nonlinear transformation further amplifies the differences in resource projections in specific directions, making the formula more sensitive to resources with significant features.

[0062] In order to more accurately reflect the dynamic market characteristics of resources, the formula also introduces the time derivative of disease distribution , which is the rate of change of resource demand. This part reflects how the market demand for resources changes over time, thereby dynamically capturing the changing characteristics of the market. The market growth rate of resources depends not only on their current market performance, but also on future market demand trends. By introducing the time derivative, the formula can predict the rate of change of resource market demand, ensuring that the calculated growth rate can better adapt to the dynamic changes of the future market. Nonlinear transformation function The use of further enhances the adaptability of the formula to complex market dynamics. It can make appropriate adjustments and changes to the distribution of disease types, making this dynamic change more in line with the actual situation. In addition, the adjustment parameters The introduction of provides the formula with the ability to adjust, so that the formula can flexibly adjust the influence of the characteristic value on the growth rate according to different application scenarios. The formula can control the weight of the influence of the characteristic value on the growth rate and adapt to different market environments and resource allocation strategies. This adjustment ability enables the formula to fully utilize the information in the main characteristic direction and avoid the interference of noise in the secondary characteristic direction. Exponential decay function The application in the formula ensures that the influence gradually weakens as the eigenvalue increases, thereby avoiding overfitting on data noise or minor features. This design makes the model more robust and adaptable in a changing medical market environment. Compared with traditional methods, this formula not only takes into account the performance of resources in the current market, but also comprehensively considers the future changing trend of resource demand. Such dual considerations ensure the foresight and rationality of the resource allocation strategy. In the process of optimizing the allocation of medical resources, this calculation method based on data-driven and dynamic analysis can significantly improve the utilization efficiency of resources and the scientific nature of allocation. Through the formula of Example 3, the system can accurately identify and predict the market growth trend of medical resources, thereby achieving the optimal resource allocation decision and improving the efficiency and quality of overall medical services.

[0063] Embodiment 4: The process of the Boston matrix construction and analysis unit calculating the market share of each medical resource according to the high-dimensional feature matrix specifically includes:

[0064] ;

[0065] in, For resources market share; is the eigenvector matrix Middle Line Elements of a column; For the eigenvalues; is the standard deviation of the distribution, which is used to adjust the impact of eigenvectors and eigenvalues ​​on market share; is the total number of patients.

[0066] Specifically, in the first part of the formula, market share The calculation uses an exponential function to evaluate the resource The core proportion of the market. The core idea of ​​this part is to use the eigenvectors and eigenvalues ​​of the high-dimensional feature matrix to quantify the comprehensive performance of resources in various feature directions. Elements in Represents resources In the feature direction This projection can be understood as the degree of performance of the resource in a certain characteristic dimension. The corresponding characteristic value It represents the importance weight of this feature direction in the entire dataset. and the corresponding eigenvalues By multiplying and squaring these products, the formula captures the total contribution of the resource in the high-dimensional feature space. The negative exponential form of the exponential function is reflects the market weight of this cumulative contribution under the assumption of normal distribution, where The standard deviation is used to adjust the impact of eigenvectors and eigenvalues ​​on market share. In this way, the formula can flexibly and dynamically adjust the market share of resources according to the distribution of eigenvectors and eigenvalues, thereby avoiding deviations caused by extreme values ​​of a single feature. This processing method ensures that the share of each resource in the market reflects its comprehensive performance in all feature directions, rather than relying solely on a specific direction. In addition, the formula also introduces a logarithmic transformation part to further correct the calculation of market share. This part reflects the resource The relative difference in performance of a resource compared to all other resources in the market. Specifically, the formula calculates the quantity of a resource and other resource quantities The sum of the absolute values ​​of the differences after nonlinear transformation. By calculating the logarithmic sum of these differences, the formula can capture the relative competitive relationship between resources. This competitive relationship is not only the performance of resources in the current market, but also a direct reflection of their potential in the future market. In this way, the formula effectively takes into account the competitive pressure and relative advantages between resources. The use of the logarithmic function further smoothes the distribution of the difference values ​​and avoids excessive fluctuations in market share due to extreme difference values. At the same time, the denominator of the logarithmic transformation The total number of patients Taking this into consideration, a standardized benchmark is provided to ensure that the relative market share between resources is still reasonably reflected regardless of how the market size changes. This dual approach combining exponential functions and logarithmic corrections provides a balanced model for calculating market share, which can accurately assess the relative importance of each resource in a complex medical market environment. Compared with simple linear calculation methods, this complex mathematical model of the present invention more accurately and comprehensively reflects the position of resources in the market, not only considering the performance of resources in various characteristic directions, but also incorporating their relative competitive position and dynamic changes in the entire market. In this way, the system can better understand and utilize the data characteristics of medical resources, optimize resource allocation strategies, and improve the overall efficiency of medical services.

[0067] Embodiment 5: The Boston matrix construction and analysis unit constructs the Boston matrix through the following formula to classify different medical resources and obtain classification results:

[0068] ;

[0069] in, For resources Position score in Boston matrix; is the base of natural logarithms; medical resources The classification results obtained for:

[0070] ;

[0071] in, and They are the upper and lower limits of the first classification threshold, respectively, used to distinguish the Boston Matrix score boundaries of different quadrants; and are the upper and lower limits of the second classification threshold, which are used to distinguish the market growth rate boundaries of different quadrants; , , and are four different preset positive integer values, representing the advantage diseases, cash diseases, problem diseases and disadvantage diseases in the Boston matrix model, satisfying: ,and .

[0072] Specifically, this formula is designed to calculate the location score of each medical resource in the Boston Matrix , thereby determining their position in the four quadrants: Advantageous diseases, cash diseases, problem diseases or disadvantageous diseases. The basis of this classification is two key indicators: market growth rate and market share . The market growth rate reflects the development potential of resources in the market, while the market share represents the proportion of resources in the current market. These two indicators are usually analyzed through linear or simple combinations, but in the present invention, the two indicators are analyzed more deeply through complex mathematical expressions and nonlinear transformations, revealing the interactive relationship between them and the comprehensive impact on resource classification. The design of the formula first adopts the idea of ​​logistic regression, combining the market growth rate and market share through nonlinear transformation to form a basic score part. This part uses exponential functions and logarithmic functions to dynamically combine the square of the market growth rate and the square root of the market share with their product. Such a combination can not only amplify the impact of high-growth resources, but also to a certain extent suppress the risks brought by over-concentrated market share. This basic score reflects the static performance of resources in the market. It is a smooth and continuous function, so that the score can be adaptively adjusted as market conditions change. On this basis, the formula further introduces sine functions and exponential decay functions to correct the score. The core of this design is to capture the volatility and competitive relationship of resources in the market. The introduction of the sine function means that there are cyclical changes or fluctuations between the market growth rate and the market share, which is particularly common in the medical market. For example, some resources may show a higher market growth rate under certain conditions, but may perform poorly in another market environment. This fluctuation is reflected by the sine function, so that the score can better capture this dynamic characteristic in the market. The exponential decay function is used to correct the difference between the growth rate and the market share. When the difference between the two is large, the score will be significantly reduced, which reflects the market uncertainty or risk that the resource may face. This part not only further adjusts the accuracy of the score by introducing the square and absolute value of the difference, but also enhances the system's sensitivity to extreme market conditions, making the classification results more reliable. Score After the calculation is completed, the system classifies according to the preset thresholds. These thresholds include high and low thresholds for scores and classification thresholds for market growth rates. Through these settings, resources can be divided into four categories: dominant diseases, cash diseases, problem diseases, and disadvantaged diseases. This classification is not only based on static scores, but also comprehensively considers the high and low market growth rates, so that the classification results can not only reflect the current market performance, but also predict future development potential. For example, dominant diseases represent those resources that have performed well in both market growth rate and market share, while cash diseases may be those resources that have a low market growth rate but still occupy a large market share. On the contrary, problem diseases and disadvantaged diseases represent those resource categories that need to be further optimized or eliminated. This classification method enables the medical resource optimization analysis system based on the Boston matrix model to respond more flexibly to complex and changing market environments through precise mathematical modeling. Compared with the traditional simple classification method, the formula of the present invention provides a more detailed market performance evaluation for each resource through multi-level and multi-dimensional analysis. This evaluation not only takes into account the current status of the resources, but also captures the potential risks and opportunities in the market through nonlinear correction and dynamic adjustment. Ultimately, the system can use these classification results to optimize resource allocation strategies and improve the overall operational efficiency and service quality of medical institutions.

[0073] Embodiment 6: The genetic algorithm optimization unit, based on the classification results, uses the genetic algorithm to optimize the allocation of medical resources, initializes the population, each individual in the population represents a resource allocation scheme, and each scheme is represented by a vector Indicates that It is a resource In the The amount distributed among individuals;

[0074] ;

[0075] in: and are the upper and lower bounds of resource allocation respectively; To generate arrive A random number between .

[0076] Specifically, the resource allocation in the formula The calculation uses a random initialization strategy: This strategy randomly initializes the allocation of each resource within a pre-set upper and lower bound, where and The minimum and maximum values ​​for the resource allocation are obtained by multiplying the value by a random number between 0 and 1. , the initial allocation of each resource is randomly distributed within the range between the upper and lower bounds. This random initialization strategy aims to ensure the diversity of the population and prevent the algorithm from falling into the local optimal solution in the initial stage. A diverse population can explore the solution space more comprehensively, thereby increasing the probability of the genetic algorithm finding the global optimal solution. Based on this initial population, the genetic algorithm continuously optimizes the resource allocation plan through a series of genetic operations, including selection, crossover, and mutation. In each generation of iteration, the system first evaluates each individual in the population according to a fitness function, which usually measures the quality of the resource allocation plan, such as the efficiency of resource utilization, cost-effectiveness, and the degree of satisfaction of the needs of different types of patients. Those individuals with higher fitness are more likely to be selected into the next generation, a process similar to the "survival of the fittest" principle in natural selection. Through the crossover operation, the system combines two individuals with high fitness in the current population to generate a new resource allocation plan. The mutation operation randomly changes the values ​​of certain elements of the individual to increase the diversity of the population and prevent the algorithm from falling into the local optimal solution. As the iteration proceeds, the population gradually tends to contain individuals with higher fitness, which means that the quality of the resource allocation plan gradually improves. When the preset number of iterations is reached or the fitness function reaches a convergence state, the genetic algorithm terminates and outputs the individual with the highest fitness in the current population as the final optimization result. This process ensures that the system can find an optimal solution for resource allocation, so that the configuration of medical resources can not only meet the needs of different patients, but also maximize utilization efficiency and cost-effectiveness.

[0077] Embodiment 7: When the genetic algorithm optimization unit uses the genetic algorithm to optimize the allocation of medical resources according to the classification results, the fitness function formula is as follows:

[0078] ;

[0079] in: For the The fitness value of each individual; For resources Boston Matrix score; For the Individual resources The amount of distribution; is the penalty coefficient, used to control over-allocation of resources; It is a regulation parameter used to adjust the impact of overall resource allocation deviation on fitness; Variance allocated for overall resources; is the total resource volume.

[0080] Specifically, the core idea of ​​the fitness function is to evaluate each individual The fitness of a resource allocation scheme in the population. The higher the fitness, the closer the allocation scheme is to the optimal one, and the more likely it is to be retained and further optimized during the evolution of the genetic algorithm. In this formula, the fitness value The calculation of not only considers the effect of resource allocation, but also includes the penalty for over-allocation of resources and the adjustment of the overall resource allocation deviation. First, the main part of the formula is calculated by the Boston matrix score of each resource. , classification results and resource allocation Boston Matrix Score Reflects the position and importance of each resource in the market, the classification results The strategic category of resources is identified (such as advantageous diseases, cash-flowing diseases, etc.). This information is used to weigh the resource allocation strategy so that resources with higher importance receive greater priority in allocation. The logarithmic function is used in the formula The resource allocation is smoothed. This nonlinear transformation can avoid extreme fluctuations in the fitness value caused by excessive or insufficient resource allocation. In addition, a sine function term is introduced into the fitness function to further adjust and correct the effect of the allocation strategy. This sine function term The correlation between the amount of resource allocation and the market growth rate is considered. Market Growth Rate Represents the growth potential of a resource in the market, multiplied by the amount of resource allocated After that, it is converted into the adjustment coefficient of cyclical fluctuations through the sine function. This design takes into account the different characteristics that resources may exhibit under different market conditions, ensuring that the allocation plan can adapt to the dynamic changes of the market. At the same time, the sine function term also contains a denominator term , which measures the difference between resource allocation and market share. When the difference between allocation and market share is large, this term will significantly reduce the fitness value, reflecting the deviation of resource allocation and possible inefficient configuration. This design ensures that resource allocation not only takes into account the market growth rate, but also matches the actual market share to avoid excessive or insufficient concentration of resources.

[0081] Another key part of the fitness function is the penalty for over-allocation of resources, which is calculated by a square term The purpose of this penalty term is to limit the extreme allocation of resources and prevent some resources from being wasted due to over-allocation. The fitness function is calculated by the square of the ratio of . The fitness function can impose a larger penalty on resources that are close to or exceed the allocation upper limit, reducing their fitness in the genetic algorithm. This penalty mechanism ensures the rationality and sustainability of the resource allocation strategy, and avoids management difficulties and waste caused by excessive concentration of resources in actual operations. The last part of the fitness function involves adjusting the overall resource allocation deviation. This part simulates the expected situation of overall resource allocation through a normal distribution function. Specifically, the design purpose of this part is to ensure that the total allocation of all resources Close to the estimated total resources By introducing a bias term , the fitness function can penalize those solutions whose total allocation deviates from the total resource amount. This adjustment mechanism uses an exponential function and variance The combination of can smooth the total allocation deviation and ensure that the total resource allocation is consistent with the overall resource budget and allocation strategy. It is used to adjust the impact of this deviation on fitness, increasing the flexibility and adaptability of the formula.

[0082] Embodiment 8: When the genetic algorithm optimization unit uses the genetic algorithm to optimize the allocation of medical resources according to the classification results, the selection probability formula of the selection operation is as follows:

[0083] ;

[0084] in, For the The probability of an individual being selected; For the The fitness of each individual; is the size of the population; the formula for the crossover operation is as follows:

[0085] ;

[0086] in, Resources for new individuals The amount of distribution; and Parent individuals and Chinese Resources The amount of distribution; is the crossover ratio, with a value of arrive between; is a random angle value used to introduce nonlinear changes; the formula for the mutation operation is as follows:

[0087] ;

[0088] in, The resources of the new individuals after mutation The amount of distribution; The resources of the new individuals generated after crossover The amount of distribution; is the coefficient of variation, controlling the amplitude of variation; is a standard normally distributed random variable; To adjust the parameters, the control variation range changes with the deviation of market share.

[0089] Specifically, the selection operation of the genetic algorithm optimization process is based on the probability formula To determine the probability of each individual being selected. This selection strategy is based on the individual's fitness value , the higher the fitness value of an individual, the greater the probability of being selected. The fitness value is a key indicator to measure the quality of a resource allocation plan. It comprehensively considers the market performance of resources, the allocation effect, and the efficiency of resource utilization. Through this fitness-based selection method, the genetic algorithm can effectively retain those individuals with excellent performance so that their genes can be passed on to the next generation. At the same time, it also retains a part of individuals with lower fitness to maintain the diversity of the population. Population diversity is one of the core advantages of genetic algorithms, which can prevent the algorithm from falling into local optimality and ensure the finding of the global optimal solution. This design of the selection operation reflects the simulation of the natural selection mechanism, that is, the evolutionary process of "survival of the fittest". After selecting individuals with higher fitness, the genetic algorithm generates a new resource allocation plan through crossover operations to further optimize the population. Formula for crossover operation Combined with parent individuals and The resource allocation amount is calculated by weighting coefficient Control the influence of the parent generation on the new individual. This weighted combination method allows the new individual to not only inherit the good characteristics of the parent generation, but also A certain degree of randomness and variability is introduced to increase the diversity of understanding. The introduction of the sine function can effectively break the limitations of simple linear combinations, allowing new individuals to have a wider range of exploration capabilities and better adapt to the complex medical resource allocation environment. The use of introduces nonlinear changes, making new individuals show greater diversity in genetic characteristics, which helps to search more efficiently in complex multidimensional solution spaces. The mutation operation is a key step in introducing random mutations in genetic algorithms, aiming to increase the diversity of the population and prevent the algorithm from falling into local optimality. The mutation operation is performed through the formula The coefficient of variation in the formula is Control the amplitude of variation to ensure the rationality and controllability of variation. Standard normal distribution random variable The introduction of makes the mutation have a certain randomness, which plays a role in "stirring" the population during the optimization process, breaking the structure of the current population and exploring new solution space. In addition, the exponential decay term in the formula The adjustment is made according to the deviation between the resource allocation and the market share. This design takes into account that when the gap between resource allocation and market share is large, greater adjustments and explorations are required, so a larger variation range is allowed. When the resource allocation is close to the market share, the variation range will be relatively small to maintain the stability of the current solution. This mutation mechanism can dynamically adjust the optimization strategy, so that the algorithm shows different optimization directions at different stages, so as to better adapt to the actual changes in demand for medical resources. Through the cooperation of the above selection, crossover and mutation operations, the genetic algorithm can gradually optimize the resource allocation scheme in the population in multiple generations of iterations and finally find the optimal solution. The optimization process of each generation takes into account the market performance and utilization efficiency of resources, and adjusts and optimizes them through complex mathematical models to ensure that the final resource allocation scheme can maximize the utilization efficiency and benefits of resources. Compared with traditional optimization methods, the advantage of genetic algorithms is that they do not depend on the specific structure of the problem, have strong robustness and adaptability, and can maintain efficient optimization performance in a complex and changing medical market environment. This method is particularly suitable for solving complex optimization problems involving multiple interrelated factors, such as the allocation and configuration of medical resources. By introducing genetic algorithms into the medical resource optimization analysis system based on the Boston Matrix model, the present invention can effectively integrate the market performance and actual demand of resources, and dynamically adjust the resource allocation strategy to cope with changes and challenges in the medical environment. Ultimately, through this intelligent and data-driven optimization method, the system can improve the efficiency of resource allocation, reduce waste, improve the overall quality of medical services and patient satisfaction, and provide a scientific and efficient solution for resource management and optimization of medical institutions. Through this optimization process, medical institutions can not only better meet the diverse needs of patients, but also achieve the best operating results under limited resources. The use of genetic algorithms enables the present invention to flexibly respond to the complexity and uncertainty of the medical market and realize the intelligence and optimization of resource allocation.

[0090] Embodiment 9: The genetic algorithm optimization unit uses the Lagrange multiplier method to constrain when continuously iterating and optimizing the allocation plan of medical resources. The formula is as follows:

[0091] ;

[0092] ;

[0093] in, is the Lagrange multiplier; Assigning vectors to resources; is the constraint function, which is expressed using the following formula:

[0094] .

[0095] Specifically, the Lagrangian function In the design of the first part, the focus is on measuring the deviation between the actual allocation of resources and the optimal allocation. , the formula takes into account the resources Actual and ideal distribution The logarithmic difference between the two, this difference measure can capture the efficiency loss of the allocation scheme. In the real medical resource allocation scenario, the optimal allocation of each resource (such as beds, drugs, equipment, etc.) is based on the precise calculation of demand forecasts and resource utilization. Therefore, any actual allocation that deviates from the optimal allocation will result in a certain efficiency loss. This loss is quantified by the logarithmic term, which can more sensitively reflect the rationality and degree of optimization of resource allocation. By optimizing this part, the system strives to make the resource allocation as close to its optimal value as possible, ensuring the maximization of resource utilization and the minimization of waste. The second part of the formula introduces a quadratic penalty term to deal with resource allocation exceeding its maximum available amount. Specifically, through , the formula imposes a quadratic penalty on resource allocations that exceed the available range. The logic behind this design idea is that over-allocation of resources will not only cause actual waste of materials, but may also cause operational problems, such as imbalanced resource allocation and delayed patient treatment. By squaring these over-allocated amounts and combining them with the Lagrange multiplier method, the system can effectively prevent the phenomenon of over-concentration of resource allocation and ensure the rationality and balance of resource allocation. In addition, the parameter The setting also increases the flexibility of the formula. It can adjust the penalty intensity of resource over-allocation according to different scenarios and needs, so as to better adapt to various complex situations in practical applications.

[0096] In the Lagrangian function, the constraint function The introduction of is to ensure that the resource allocation scheme can be optimized under certain constraints. For medical resource allocation, the most common constraints include total resource limit, maximum and minimum available amount of a single resource, etc. By defining the constraint function , the system can accurately describe these constraints. For example, Used to ensure that the total amount of all allocated resources does not exceed the specified total amount ; and They are used to restrict the allocation of a single resource to the maximum and minimum available amounts. These constraints are assigned Lagrange multipliers , these multipliers can be understood as the influence of constraints on the objective function. By introducing these Lagrangian multipliers, the system can dynamically adjust resource allocation during the optimization process to ensure that all constraints are met. A key advantage of the Lagrangian multiplier method is that it can handle multiple constraints at the same time without explicitly including these constraints in every step of the optimization process. Instead, by introducing Lagrangian multipliers, the constraints are converted into part of the optimization objective, making the entire optimization process more coherent and efficient. By solving the partial derivatives of the Lagrangian function and setting them to zero, the optimal value of each resource allocation can be obtained. This process not only ensures that the solution found satisfies all constraints, but also maximizes the value of the objective function to achieve the optimal effect of resource allocation. The application of this optimization method in the allocation of medical resources has significant practical significance. The allocation of medical resources often needs to be optimized under complex constraints. For example, in the emergency department, the number of beds is limited and needs to be dynamically adjusted according to the severity of the patient's condition. In this case, the use of the Lagrangian multiplier method can ensure the reasonable allocation of key resources such as beds while meeting the needs of all patients. The Lagrange multiplier method can also adapt to the uncertainty of resource conditions and demands, and by dynamically adjusting resource allocation strategies, the system can respond quickly to changing medical needs.

[0097] The present invention is described in detail above. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. It should be pointed out that for ordinary technicians in this technical field, without departing from the principle of the present invention, several improvements and modifications can be made to the present invention, and these improvements and modifications also fall within the scope of protection of the claims of the present invention.

Claims

1. A medical resource optimization analysis system based on the Boston Matrix model, characterized by: The system includes: a data acquisition and preprocessing unit, a Boston matrix construction and analysis unit and a genetic algorithm optimization unit; the data acquisition and processing unit is used to collect medical data from various data sources, delete redundant and erroneous data in the medical data, fill in missing values, extract features after nonlinear transformation of the medical data, and then use a nonlinear mapping function to perform high-dimensional mapping on the features to generate a high-dimensional feature matrix; the Boston matrix construction and analysis unit is used to calculate the market growth rate and market share of each medical resource according to the high-dimensional feature matrix, construct a Boston matrix, classify different medical resources, and obtain classification results; the genetic algorithm optimization unit uses a genetic algorithm to optimize the allocation of medical resources according to the classification results, and continuously iterates and optimizes the allocation plan of medical resources through selection, crossover and mutation operations. When the iterative optimization reaches a set number of times, the final resource allocation vector is output as the optimal configuration of medical resources; the data acquisition and preprocessing unit includes: a data acquisition module and a preprocessing module; the data acquisition module collects multi-dimensional medical data to form a data set to form a data set: X={X1,X2,…,X n }, Each data point X i The data point corresponding to medical resource i, n is the total number of medical resources, including the following attributes: Number of patients N i , represents the number of patients treated in a given period of time and the utilization rate of medical equipment U i , indicating the frequency of use of medical equipment and the distribution of diseases in a given period of time D i , which represents the proportion of patients with different diseases and other related variables in a given period of time i ; φ i Including: bed occupancy rate, drug usage, doctor working hours and / or nurse working hours; the preprocessing unit is used to delete redundant and erroneous data in the data set, fill in missing values, extract features after nonlinear transformation of the data set, and then use nonlinear mapping function to perform high-dimensional mapping on the features to generate a high-dimensional feature matrix M; the Boston matrix construction and analysis unit calculates the market growth rate of each medical resource according to the high-dimensional feature matrix, which specifically includes: eigenvalue decomposition of the high-dimensional feature matrix M: M=QΛQ -1 ; Among them, Q is the eigenvector matrix; Λ is the diagonal eigenvalue matrix; using the eigenvalues ​​and eigenvectors in the eigenvalue decomposition results, combined with the dynamic characteristics of resources, the growth rate calculation formula is constructed: Among them, G i is the market growth rate of resource i; n is the total number of resources; λ k is the kth eigenvalue, indicating the contribution of resources in a specific characteristic direction; q ik is the element in the i-th row and k-th column of the feature vector matrix Q, representing the projection of resource i in the k-th feature direction; is the time derivative of the disease distribution of resource i, indicating the speed of change of disease distribution, reflecting the dynamic characteristics of market growth; T(·) represents a nonlinear transformation function; μ is an adjustment parameter used to control the influence of the eigenvalue on the growth rate, which is a set value; the Boston matrix construction and analysis unit calculates the market share of each medical resource according to the high-dimensional feature matrix, specifically including: Among them, S i is the market share of resource i; q ij is the element in the i-th row and j-th column of the eigenvector matrix Q; j is the jth eigenvalue; σ is the standard deviation of the distribution, which is used to adjust the impact of eigenvectors and eigenvalues ​​on market share; N total is the total number of patients; the Boston matrix construction and analysis unit constructs the Boston matrix through the following formula, classifies different medical resources, and obtains the classification result: Among them, B i is the position score of resource i in the Boston matrix; e is the base of the natural logarithm; the classification result Class(i) obtained by medical resource i is: where τ high and τ low are respectively the upper and lower limits of the set first classification threshold, used to distinguish the score boundaries of the Boston matrix in different quadrants; θ high and θ low are respectively the upper and lower limits of the set second classification threshold, used to distinguish the market growth rate boundaries in different quadrants; where A, B, C, and D are respectively four preset different positive integer values, representing the dominant disease types, cash disease types, problem disease types, and disadvantaged disease types in the Boston matrix model, satisfying: A > B > C > D, and 4 < A + B + C + D < 10; when the genetic algorithm optimization unit optimizes the allocation of medical resources according to the classification results, it initializes the population, and each individual in the population represents a resource allocation plan, and each plan is represented by a vector indicating that is the allocation amount of resource i in the k-th individual; Where: R min and R max are the upper and lower bounds of resource allocation respectively; rand() is used to generate a random number between 0 and 1; when the genetic algorithm optimization unit uses the genetic algorithm to optimize the allocation of medical resources according to the classification results, the formula of the fitness function is as follows: Where: F(R k ) is the fitness value of the kth individual; B i is the Boston matrix score of resource i; is the allocation of resource i in the kth individual; λ is the penalty coefficient, which is used to control over-allocation of resources; η is the adjustment parameter, which is used to adjust the impact of the overall resource allocation deviation on fitness; U 2 is the variance of overall resource allocation; R total is the total amount of resources; when the genetic algorithm optimization unit uses the genetic algorithm to optimize the allocation of medical resources according to the classification results, the selection probability formula of the selection operation is as follows: Among them, P(R k ) is the probability that the kth individual is selected; F(R k ) is the fitness of the kth individual; N pop is the size of the population; the formula for the crossover operation is as follows: in, is the allocation of resource i in the new individual; and are the allocation amounts of resource i in the parent individuals p and q respectively; α is the crossover ratio, which takes values ​​between 0 and 1; θ is the random angle value, which is used to introduce nonlinear changes; the formula for the mutation operation is as follows: in, is the allocation amount of resource i in the new individual after mutation; is the allocation of resource i in the new individuals generated after crossover; δ is the coefficient of variation, which controls the amplitude of variation; is a standard normally distributed random variable; ρ is an adjustment parameter that controls the variation of the amplitude with the deviation of market share.

2. The medical resource optimization analysis system based on the Boston matrix model according to claim 1, characterized in that: The genetic algorithm optimization unit uses the Lagrange multiplier method to constrain when continuously iterating and optimizing the allocation plan of medical resources. The formula is as follows: Where L=(L1,L2,…,L m ) is the Lagrange multiplier; R is the resource allocation vector; g j (R) is the constraint function, which is expressed by the following formula:

Images