A Clinical Intelligent Scheduling Method and System Based on Big Data Analysis

Abstract

The present invention relates to a clinical intelligent scheduling method and system based on big data analysis, belonging to the technical field of data processing. The method includes: obtaining a data set of patient data, nurse data, and environmental data, and preprocessing the data set to generate a standardized data set; processing the standardized data set to predict the nursing needs for future scheduling; generating a hierarchical nurse allocation plan through a recommendation algorithm based on the skills of nurses, the current workload, and historical work data; using a genetic algorithm combined with dynamic constraint conditions to generate a work schedule that meets the hierarchical nurse allocation plan and nursing needs; dynamically adjusting the workload of each nurse based on the scheduling data and physiological indicators of the work schedule through a supervised learning model to generate a first scheduling strategy; automatically matching substitute nurses and optimizing the first scheduling strategy in case of emergencies through a reinforcement learning model to generate a second scheduling strategy. The present invention can improve the efficiency and accuracy of clinical scheduling.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and particularly to a clinical intelligent scheduling method, system, electronic device, and non-transitory computer-readable storage medium based on big data analysis. Background Art

[0002] Currently, clinical nursing scheduling mainly relies on manual experience for deployment. The head nurse needs to comprehensively consider the balance of working hours, skill matching, and possible emergencies to ensure the smooth progress of nursing work.

[0003] However, this method highly depends on personal experience, not only with a large workload and low efficiency, but also prone to unreasonable scheduling due to human negligence, affecting the quality of nursing. In addition, the traditional scheduling method mostly adopts a fixed shift system, which is difficult to flexibly adapt to the fluctuations in patient needs, such as the increased need for night care for postoperative patients or the centralized arrangement of rehabilitation training, easily resulting in waste or shortage of nursing resources. Summary of the Invention

[0004] The present invention aims at the technical problems existing in the prior art, and provides a clinical intelligent scheduling method, system, electronic device, and non-transitory computer-readable storage medium based on big data analysis that can improve the scheduling efficiency and accuracy.

[0005] The technical solution for the present invention to solve the above technical problems is as follows:

[0006] The present invention provides a clinical intelligent scheduling method based on big data analysis, and the method includes:

[0007] Obtain a data set of patient data, nurse data, and environmental data, and preprocess the data set to generate a standardized data set;

[0008] Use a time series prediction model and a decision tree model to process the standardized data set to predict the nursing needs for future scheduling;

[0009] Based on the skills, current workload, and historical work data of nurses, generate a hierarchical nurse allocation plan through a recommendation algorithm;

[0010] Adopt a genetic algorithm combined with dynamic constraint conditions to generate a schedule that meets the hierarchical nurse allocation plan and the nursing needs;

[0011] Based on the scheduling data and physiological indicators of the schedule, dynamically adjust the workload of each nurse through a supervised learning model to generate a first scheduling strategy;

[0012] Automatically match substitute nurses and optimize the first scheduling strategy in case of emergencies through a reinforcement learning model to generate a second scheduling strategy.

[0013] Optionally, the preprocessing of the data set to generate a standardized data set includes:

[0014] Processing the nurse data using the K-means clustering algorithm, grouping according to the skill level and work intensity of the nurses, and obtaining the grouped nurse data;

[0015] Using the PCA dimensionality reduction algorithm to reduce the dimensionality of the multi-dimensional feature data of the grouped nurse data, the patient data, and the environmental data and extract key features to obtain the standardized data set.

[0016] Optionally, the time series prediction model is an LSTM network, and the decision tree model is an XGBoost algorithm;

[0017] Processing the standardized data set using the time series prediction model and the decision tree model to predict the nursing needs for future shifts, including:

[0018] Using the LSTM network to process the standardized data set to predict the nursing demand fluctuations of postoperative patients at night;

[0019] Using the XGBoost algorithm to process the standardized data set to predict the nursing demand level for future shifts;

[0020] Determining the nursing needs for future shifts according to the nursing demand fluctuations and the nursing demand level.

[0021] Optionally, generating a hierarchical nurse allocation plan based on the skills of the nurses, the current workload, and historical work data, including:

[0022] The collaborative filtering recommendation algorithm matches the fitness of each nurse with the shift schedule based on historical shift data;

[0023] The KNN algorithm calculates the workload similarity between nurses;

[0024] Generating the hierarchical nurse allocation plan according to the fitness and the workload similarity.

[0025] Optionally, the dynamic constraint conditions of the genetic algorithm include daily working hours constraint, weekly shift days constraint, and consecutive rest days constraint.

[0026] Optionally, the matching of the fitness of each nurse with the shift schedule based on historical shift data includes:

[0027] Constructing a first fitness function for maximizing the weighted sum of the nurse satisfaction score and the nursing demand coverage rate;

[0028] Construct a second fitness function for minimizing the product of the sum of nurse fatigue indices and the number of scheduling adjustments;

[0029] Determine the fitness according to the function value of the first fitness function and the function value of the second fitness function.

[0030] Optionally, the physiological index includes heart rate variability data, and the supervised learning model is a random forest classifier;

[0031] Based on the scheduling data and physiological indices of the scheduling table, dynamically adjust the workload of each nurse through a supervised learning model to generate a first scheduling strategy, including:

[0032] Use the random forest classifier to calculate the fatigue index of the nurse in combination with the scheduling data and the nurse's heart rate variability data;

[0033] Dynamically adjust the workload of each nurse according to the fatigue index of the nurse to generate the first scheduling strategy.

[0034] Optionally, the reinforcement learning model is a Deep Q-Learning model; Automatically match substitute nurses and optimize the first scheduling strategy in case of emergencies through the reinforcement learning model to generate a second scheduling strategy, including:

[0035] Obtain the linear combination of the skill matching degree, fatigue index and scheduling time conflict rate of each substitute nurse;

[0036] Optimize the first scheduling strategy according to the linear combination of the skill matching degree, fatigue index and scheduling time conflict rate of each substitute nurse, and the optimization goal is to minimize the decline in nursing quality and the additional workload of nurses in the current scheduling to generate the second scheduling strategy.

[0037] Optionally, the nurse data includes the nurse's skill level, scheduling preference, night shift acceptance flag and historical work log, and the environmental data includes the predicted increase in nursing demand during the peak season of seasonal diseases.

[0038] In addition, to achieve the above object, the present invention also proposes a clinical intelligent scheduling system based on big data analysis, and the system includes:

[0039] A data acquisition module for acquiring a data set of patient data, nurse data and environmental data, and preprocessing the data set to generate a standardized data set;

[0040] A demand prediction module for using a time series prediction model and a decision tree model to process the standardized data set to predict the nursing demand for future scheduling;

[0041] A hierarchical allocation module, which is used to generate a hierarchical nurse allocation plan through a recommendation algorithm based on the skills, current workload and historical work data of nurses;

[0042] A preliminary scheduling module, which is used to generate a schedule that meets the hierarchical nurse allocation plan and the nursing requirements by using a genetic algorithm combined with dynamic constraint conditions;

[0043] A first scheduling module, which is used to dynamically adjust the workload of each nurse based on the scheduling data and physiological indicators of the schedule through a supervised learning model, and generate a first scheduling strategy;

[0044] A second scheduling module, which is used to automatically match substitute nurses and optimize the first scheduling strategy in case of emergencies through a reinforcement learning model, and generate a second scheduling strategy.

[0045] In addition, to achieve the above object, the present invention also provides an electronic device, including: a memory for storing a computer software program; a processor for reading and executing the computer software program, thereby implementing a clinical intelligent scheduling method based on big data analysis as described above.

[0046] In addition, to achieve the above object, the present invention also provides a non-transitory computer-readable storage medium, in which a computer software program is stored, and when the computer software program is executed by a processor, a clinical intelligent scheduling method based on big data analysis as described above is implemented.

[0047] The beneficial effects of the present invention are:

[0048] (1) By intelligently matching nurse resources, the present invention assigns suitable nurses to suitable patient care tasks according to the skills, experience and workload of nurses, ensuring that patients can receive the most professional nursing services. It can predict the nursing requirements for a period of time in the future based on historical data and real-time data, dynamically adjust the number of nurses, avoid waste or shortage of nursing resources, and ensure that the nursing needs of patients can be met at different times.

[0049] (2) The present invention generates an optimal schedule through a genetic algorithm, taking into account the work preferences, fatigue index and working hour constraints of nurses, ensuring that the workload of each nurse is balanced, avoiding overwork of individual nurses, and at the same time trying to meet the rest needs of nurses. The goal of the genetic algorithm is "the least work conflict and the least adjustment cost". By dynamically adjusting the number of nurses and combining the results of nursing demand prediction, it can effectively reduce work conflicts caused by unreasonable scheduling and improve the job satisfaction of nurses.

[0050] (3) The present invention can automatically calculate the working hours and the number of night shifts of nurses, automatically identify legal holidays, nurses' birthdays, and nurses' scheduled vacation times, and optimize attendance management. By using a random forest classification model to predict whether nurses need to adjust their shifts due to fatigue or health reasons, the workload of manual attendance management is reduced, and the management efficiency and accuracy are improved.

[0051] In summary, through an intelligent and data-driven approach, the present invention realizes the efficient management and optimized scheduling of nursing resources, not only improving the quality and efficiency of nursing care, but also enhancing the professional satisfaction of nurses and patient satisfaction. At the same time, it reduces management costs and strengthens the overall operation ability of the hospital. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 It is a scenario diagram of a clinical intelligent scheduling method based on big data analysis provided by the present invention;

[0053] Figure 2 It is a flowchart of a clinical intelligent scheduling method based on big data analysis provided by the present invention;

[0054] Figure 3 It is a schematic structural diagram of a clinical intelligent scheduling system based on big data analysis provided by the present invention;

[0055] Figure 4 It is a schematic hardware structure diagram of a possible electronic device provided by the present invention;

[0056] Figure 5 It is a schematic hardware structure diagram of a possible computer-readable storage medium provided by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0057] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts fall within the protection scope of the present invention.

[0058] In the description of the present invention, the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more of the described features. In the description of the present invention, "a plurality" means two or more, unless otherwise specifically defined.

[0059] In the description of the present invention, the term "for example" is used to mean "serving as an example, illustration, or explanation". Any embodiment described as "for example" in the present invention is not necessarily construed as being more preferred or advantageous than other embodiments. The following description is given to enable any person skilled in the art to implement and use the present invention. In the following description, details are set forth for purposes of explanation. It should be understood that those of ordinary skill in the art can recognize that the present invention can be implemented without these specific details. In other instances, well-known structures and processes are not elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but rather is consistent with the broadest scope that conforms to the principles and features disclosed in the present invention.

[0060] Please refer to Figure 1 , Figure 1 which is a scenario diagram of a clinical intelligent scheduling method based on big data analysis provided by the present invention. As Figure 1 shown, the terminal and the server are connected through a network, for example, through a wired or wireless network connection, etc. Among them, the terminal may include, but is not limited to, portable terminals such as mobile phones and tablets installed with various network platform applications, as well as fixed terminals such as computers, inquiry machines, and advertising machines. Among them, the server provides various business services for users, including service push servers, user recommendation servers, etc.

[0061] It should be noted that Figure 1 the scenario diagram of a clinical intelligent scheduling method based on big data analysis shown is only an example. The terminal, server, and application scenarios described in the embodiments of the present invention are for more clearly explaining the technical solutions of the embodiments of the present invention, and do not generate limitations on the technical solutions provided by the embodiments of the present invention. Those of ordinary skill in the art can know that with the evolution of the system and the emergence of new business scenarios, the technical solutions provided by the embodiments of the present invention are equally applicable to similar technical problems.

[0062] Among them, the terminal can be used for:

[0063] Obtaining a data set of patient data, nurse data, and environmental data, and preprocessing the data set to generate a standardized data set;

[0064] Using a time series prediction model and a decision tree model to process the standardized data set to predict the nursing needs for future scheduling;

[0065] Based on the skills of nurses, current workload, and historical work data, generating a hierarchical nurse allocation plan through a recommendation algorithm;

[0066] Adopting a genetic algorithm combined with dynamic constraint conditions to generate a schedule that meets the hierarchical nurse allocation plan and the nursing needs;

[0067] Based on the shift scheduling data and physiological indicators in the shift schedule, dynamically adjust the workloads of each nurse through a supervised learning model to generate a first shift scheduling strategy;

[0068] Automatically match substitute nurses in case of emergencies through a reinforcement learning model and optimize the first shift scheduling strategy to generate a second shift scheduling strategy.

[0069] Please refer to Figure 2 , which provides a flowchart of a clinical intelligent shift scheduling method based on big data analysis according to the present invention, including the following steps:

[0070] Step 201, obtain a data set of patient data, nurse data, and environmental data, and preprocess the data set to generate a standardized data set.

[0071] In some embodiments, step 201 may include:

[0072] Use the K-means clustering algorithm to process the nurse data, group them according to the skill levels and work intensities of the nurses, and obtain the grouped nurse data;

[0073] Use the PCA dimensionality reduction algorithm to reduce the dimensionality of the multi-dimensional feature data of the grouped nurse data, the patient data, and the environmental data and extract key features to obtain the standardized data set.

[0074] Among them, the nurse data may include the skill levels of nurses, shift preferences, night shift acceptance flags, and historical work logs, and the environmental data may include the incremental nursing needs predicted during the peak season of seasonal diseases.

[0075] Specifically, the K-means clustering algorithm can be used to group the nurse data, classify them according to the skill levels, work experience, work intensities, etc. of the nurses, ensure the reasonable allocation of nurses at all levels during shift scheduling, and achieve the efficient cooperation of the nursing team.

[0076] The input data may include:

[0077] Nurse skill levels: such as junior nurses, intermediate nurses, senior nurses, etc., reflecting the professional technical levels of nurses.

[0078] Work experience: The working years of nurses, reflecting their accumulation and experience in nursing work.

[0079] Work intensity: Calculated comprehensively through working hours, night shift frequencies, the number of patients nursed, etc., reflecting the current workloads of nurses.

[0080] Specifically, K nurse data points can be randomly selected as the initial clustering centers. The determination of the value of K can be based on actual needs and experience. For example, nurses can be divided into 3 groups (junior, intermediate, senior) or more groups. Calculate the distance from each nurse data point to each clustering center, and assign the nurse to the group represented by the nearest clustering center. The distance is usually calculated using the Euclidean distance. Then recalculate each clustering center, and take the mean of all nurse data points within each group as the new clustering center. Repeat the assignment stage and the update stage until the clustering centers no longer change or reach the set number of iterations.

[0081] The output result is the grouped nurse data. Nurses are divided into different groups, and nurses within each group are similar in terms of skill level, work experience, work intensity, etc. For example, nurses in the junior nurse group may have a lower skill level, less work experience, but a relatively higher work intensity; nurses in the senior nurse group have a high skill level, rich experience, and a moderate work intensity.

[0082] The PCA dimensionality reduction algorithm reduces the dimensionality of the multi-dimensional feature data of the grouped nurse data, patient data, and environmental data, extracts key features, reduces the data dimension, improves the subsequent calculation efficiency, and at the same time retains the main information of the data.

[0083] The input data can include:

[0084] The grouped nurse data: including features such as the skill level, work experience, and work intensity of nurses.

[0085] Patient data: such as fracture type, surgical method, postoperative nursing needs, real-time vital signs, etc.

[0086] Environmental data: such as national legal holidays, hospital working day arrangements, etc.

[0087] Specifically, the input data can be standardized first so that the mean of each feature is 0 and the variance is 1. Then calculate the covariance matrix, which reflects the correlation between features. Then solve the eigenvalues and eigenvectors of the covariance matrix. The eigenvalues represent the importance of each principal component, and the eigenvectors represent the directions of the principal components. Sort the eigenvalues from largest to smallest, and select the first k eigenvectors as the principal components. Finally, project the original data onto the principal component directions to obtain the dimensionality-reduced data.

[0088] The output result can be a standardized data set and a dimensionality-reduced data set, which retains the main information of the original data but reduces the dimension and the computational complexity. For example, the original data may have dozens of features, and after PCA dimensionality reduction, there may only be a few key features left, and these features can better reflect the internal structure and laws of the data.

[0089] In summary, the present invention groups nurse data through the K-means clustering algorithm, achieving reasonable classification of nurses and providing a basis for subsequent shift scheduling. The PCA dimensionality reduction algorithm then performs dimensionality reduction processing on the grouped nurse data, patient data, and environmental data, extracts key features, improves data processing efficiency, and provides a more concise and efficient data basis for subsequent machine learning model training and optimization.

[0090] Step 202: Use a time series prediction model and a decision tree model to process the standardized data set to predict the nursing needs for future shift scheduling.

[0091] In some embodiments, the time series prediction model is an LSTM network, and the decision tree model is an XGBoost algorithm. Step 202 may include:

[0092] Use the LSTM network to process the standardized data set to predict the fluctuation of nursing needs for postoperative patients at night;

[0093] Use the XGBoost algorithm to process the standardized data set to predict the level of nursing needs for future shift scheduling;

[0094] Determine the nursing needs for future shift scheduling based on the fluctuation of nursing needs and the level of nursing needs.

[0095] In specific implementation, the LSTM network and the XGBoost algorithm are respectively used to predict the fluctuation of nursing needs for postoperative patients at night and the level of nursing needs for future shift scheduling.

[0096] Predict the fluctuation of nursing needs for postoperative patients at night through the LSTM network to arrange sufficient nursing resources in advance. The input data is a standardized data set, a data set after PCA dimensionality reduction processing, including time series features such as the historical nursing need data of patients, real-time vital signs, and the changing trend of postoperative nursing needs.

[0097] LSTM (Long Short-Term Memory network) is a special recurrent neural network (RNN) dedicated to processing time series data and capable of capturing long-term dependencies. Its core is to control the flow of information through a series of gates (input gate, forget gate, and output gate), avoiding the gradient vanishing problem of traditional RNNs.

[0098] In the data preparation stage, time series data (such as historical data on the nocturnal nursing needs of postoperative patients) can be divided into a training set and a test set. The data is converted into the format required by LSTM, usually a three-dimensional array (number of samples, number of time steps, number of features). An LSTM network is constructed, including an input layer, an LSTM layer, a fully connected layer, and an output layer. Appropriate activation functions (such as ReLU or Sigmoid) and optimizers (such as Adam) are selected. During the training process, the training set data is used to train the LSTM model, and the network parameters are updated through backpropagation to minimize the prediction error. During the training process, a validation set is used for hyperparameter tuning, such as the learning rate, the number of hidden units, the number of time steps, etc. The trained LSTM model can be used to predict the test set data to obtain the predicted values of the nocturnal nursing needs of postoperative patients. The performance of the model is evaluated, such as calculating metrics like the mean squared error (MSE) and the mean absolute error (MAE).

[0099] The output result is the prediction of the fluctuation of nursing needs. The LSTM model outputs the predicted values of the nocturnal nursing needs of postoperative patients, reflecting the changing trends of nursing needs at different time points. For example, predicting the number of nursing times and the nursing duration required on a certain night, etc.

[0100] In the present invention, the XGBoost algorithm is used to predict the nursing need levels for future scheduling. The XGBoost algorithm predicts the nursing need levels for future scheduling (such as level 1, level 2, and level 3 nursing) in order to reasonably allocate nursing resources.

[0101] Specifically, the input data is a standardized data set, a data set after PCA dimensionality reduction processing, which includes features such as patient medical record data (such as fracture type, surgical method), postoperative nursing needs, and changes in the condition of inpatients.

[0102] XGBoost (eXtreme Gradient Boosting) is an ensemble learning algorithm based on gradient boosting, which constructs a strong learner by combining multiple weak learners (usually decision trees). It has high computational performance and good prediction effects.

[0103] In the data preparation stage, the data set can be divided into a training set and a test set. The data is preprocessed, such as feature encoding, standardization, etc. An XGBoost model is constructed, and hyperparameters are set, such as the depth of the tree, the learning rate, the regularization parameter, etc. The training set data is used to train the model, and the model performance is gradually optimized through the gradient boosting algorithm.

[0104] During the training process, XGBoost will automatically handle issues such as missing values and feature selection to improve the robustness of the model. Methods such as cross-validation are used to evaluate the model performance, and hyperparameters are adjusted to obtain the best model. Then, the trained XGBoost model is used to predict the test set data to obtain the nursing requirement level for each patient. The model performance is evaluated, such as calculating metrics like accuracy, recall rate, F1 score, etc.

[0105] The output result can be the prediction of the nursing requirement level. The XGBoost model outputs the nursing requirement level for each patient (such as first-level, second-level, and third-level nursing). For example, it is predicted that a certain patient needs moderate care and another patient needs severe care.

[0106] It can be understood that determining the nursing requirements for future scheduling is based on the nursing requirement fluctuations predicted by LSTM and the nursing requirement levels predicted by XGBoost, and comprehensively determines the nursing requirements for future scheduling to reasonably arrange nurse resources.

[0107] In the data integration stage, the nursing requirement fluctuations predicted by LSTM at night and the nursing requirement levels predicted by XGBoost can be integrated. For example, for a certain night, LSTM predicts that 10 nursing operations are needed, and XGBoost predicts that among them, 5 patients need second-level nursing and 3 patients need first-level nursing. The total nursing requirements can be calculated according to the nursing requirement level and fluctuations. For example, assume that one nurse is needed for third-level nursing, two nurses are needed for second-level nursing, and three nurses are needed for first-level nursing. If the number of nursing operations predicted by LSTM is 10 times, and the distribution of the nursing requirement levels predicted by XGBoost is: third-level nursing: 2 times, second-level nursing: 5 times, first-level nursing: 3 times, then the total nursing requirements are: 2 + 10 + 9 = 21 (times).

[0108] Then, according to the calculated nursing requirements, combined with the skills, experience, and workload of the nurses, through intelligent matching of nurse resources and flexible scheduling optimization, an optimal scheduling table can be generated. For example, according to the nursing requirement level and fluctuations, reasonably arrange the combination of senior nurses and junior nurses to ensure the efficient progress of nursing work.

[0109] In summary, through the combination of the LSTM network and the XGBoost algorithm, the present invention can accurately predict the nursing requirement fluctuations at night for postoperative patients and the nursing requirement levels for future scheduling. It not only considers the changing trend of the time series but also combines the individual characteristics and condition changes of the patients, providing a scientific basis for reasonably arranging nursing resources. Finally, through intelligent matching and scheduling optimization, it ensures the efficiency, fairness, and quality of nursing work.

[0110] Step 203: Generate a hierarchical nurse allocation plan based on the skills of the nurses, the current workload, and the historical work data through a recommendation algorithm.

[0111] In some embodiments, step 203 may include:

[0112] Adopt a collaborative filtering recommendation algorithm to match the fitness of each nurse and the scheduling shift based on historical scheduling data;

[0113] Adopt the KNN algorithm to calculate the workload similarity among nurses;

[0114] Generate the hierarchical nurse allocation plan according to the fitness and the workload similarity.

[0115] Specifically, the collaborative filtering recommendation algorithm and the KNN algorithm are respectively used to evaluate the fitness of nurses and the scheduling shifts and the workload similarity among nurses, and then generate a hierarchical nurse allocation plan. The collaborative filtering recommendation algorithm is based on historical scheduling data to evaluate the fitness of each nurse and the scheduling shift, and predict the adaptability of nurses to specific shifts.

[0116] The input data is historical scheduling data, including information such as nurses' past scheduling records, nursing task completion situations, and patient feedback. The collaborative filtering recommendation algorithm is mainly divided into two categories, user-based collaborative filtering and item-based collaborative filtering. In the present invention, a user-based collaborative filtering method can be adopted.

[0117] Specifically, the similarity between nurses can be calculated, usually using cosine similarity or Pearson correlation coefficient. For example, for nurse A and nurse B, calculate the similarity of their performances (such as nursing task completion situations, patient satisfaction, etc.) in past scheduling.

[0118] Fitness prediction can predict the fitness of the current nurse for a certain shift according to the historical scheduling data of similar nurses. For example, if nurse A has taken care of postoperative fracture patients many times in the past and has performed well, the system will consider that nurse A has a higher fitness for similar shifts.

[0119] The output result can be a fitness score. The fitness scores of each nurse for different shifts reflect the adaptability of nurses to specific shifts. For example, the fitness score of nurse A for the night shift is 0.8, indicating that she has a higher adaptability to the night shift.

[0120] The purpose of the KNN algorithm is to calculate the workload similarity among nurses to ensure the balance of scheduling.

[0121] The input data can be nurses' workload data, including working hours, night shift frequency, number of nursing tasks, fatigue index, etc.

[0122] The KNN (K-Nearest Neighbors) algorithm finds the nearest neighbors by calculating the distances between data points. In this solution, the Euclidean distance can be used to calculate the workload similarity among nurses.

[0123] First, the workload data of nurses can be standardized to make different features comparable. Then, calculate the Euclidean distance between each nurse and other nurses. For example, for Nurse A and Nurse B, calculate the distances in features such as working hours and night shift frequencies. Then, based on the distance calculation results, evaluate the workload similarity among nurses. The smaller the distance, the higher the similarity.

[0124] The output result can be a similarity matrix: a matrix representing the workload similarity between each nurse and other nurses. For example, the workload similarity between Nurse A and Nurse B is 0.7, indicating that their workloads are relatively similar. Generate a hierarchical nurse allocation plan, aiming to generate a reasonable hierarchical nurse allocation plan based on the fitness of nurses to shifts and the workload similarity among nurses.

[0125] Specifically, the fitness score obtained from the collaborative filtering recommendation algorithm and the workload similarity obtained from the KNN algorithm can be combined to comprehensively evaluate each nurse. For example, for a certain shift, nurses with high fitness and low workload similarity are preferentially selected. Then, hierarchical allocation is carried out according to the skill levels and work experience of nurses. For example, senior nurses are preferentially assigned to shifts that require high skills, and junior nurses are assigned to auxiliary shifts. Ensure that the nurse combination for each shift is complementary in skills and experience while having a balanced workload. Finally, based on the comprehensive evaluation results, generate a preliminary scheduling plan. Optimize and adjust the scheduling plan to ensure compliance with constraint conditions (such as daily working hours, night shift frequencies, etc.).

[0126] The output result can be a hierarchical nurse allocation plan, such as a detailed scheduling table that clearly shows the allocation of each nurse to different shifts, ensuring the efficient, fair, and quality of nursing work.

[0127] In summary, through the combination of the collaborative filtering recommendation algorithm and the KNN algorithm, the present invention can comprehensively consider the fitness of nurses to shifts and the workload similarity among nurses, and generate a reasonable hierarchical nurse allocation plan. This plan not only improves the job satisfaction of nurses but also ensures the efficient utilization of nursing resources and the improvement of nursing quality.

[0128] In some embodiments, the fitness is determined in the following manner:

[0129] Construct a first fitness function for maximizing the weighted sum of the nurse satisfaction score and the nursing demand coverage rate;

[0130] Construct a second fitness function for minimizing the product of the sum of nurse fatigue indices and the number of schedule adjustments;

[0131] Determine the fitness based on the function value of the first fitness function and the function value of the second fitness function.

[0132] Specifically, to achieve schedule optimization, two fitness functions need to be constructed: the first fitness function and the second fitness function. These two functions are used to evaluate the performance of the schedule plan in terms of nurse satisfaction and nursing demand coverage, as well as in terms of nurse fatigue index and the number of schedule adjustments.

[0133] The purpose of the first fitness function is to construct a fitness function for maximizing the weighted sum of nurse satisfaction scores and nursing demand coverage.

[0134] The input data of the first fitness function can be the nurse satisfaction score: the nurse satisfaction score calculated based on the nurse's schedule preference, working hours, night shift frequency, etc. And the nursing demand coverage: the nursing demand coverage calculated according to the schedule plan, which reflects the degree to which the schedule plan meets the nursing demands.

[0135] Specifically, the weights of the nurse satisfaction score and the nursing demand coverage can be set, and then the weighted sum of the nurse satisfaction score and the nursing demand coverage can be calculated as the value of the first fitness function. The output result can be the value of the first fitness function, indicating the comprehensive performance of the schedule plan in terms of nurse satisfaction and nursing demand coverage.

[0136] The second fitness function can be constructed as a fitness function for minimizing the product of the sum of nurse fatigue indices and the number of schedule adjustments. The input data can be the nurse fatigue index: the nurse fatigue index calculated based on the nurse's working hours, night shift frequency, rest time, etc. And the number of schedule adjustments: the number of schedule adjustments calculated according to the schedule plan, which reflects the stability of the schedule plan.

[0137] Specifically, the product of the sum of nurse fatigue indices and the number of schedule adjustments can be calculated as the value of the second fitness function. The output result can be the value of the second fitness function, indicating the comprehensive performance of the schedule plan in terms of nurse fatigue index and the number of schedule adjustments.

[0138] The purpose of determining the final fitness is to determine the final fitness of the shift scheduling plan based on the function values of the first fitness function and the second fitness function. The function values of the first fitness function and the second fitness function can be comprehensively evaluated. Weighted sum, product or other combination methods can be used and selected according to actual needs and preferences. Then, according to the comprehensive evaluation result, the final fitness of the shift scheduling plan is determined. For example, the weighted sum of the first fitness function value and the second fitness function value can be used as the final fitness. The output result can be the final fitness, representing the comprehensive performance of the shift scheduling plan, which is used to guide the optimization and selection of the shift scheduling plan.

[0139] In summary, by constructing the first fitness function and the second fitness function, the present invention can comprehensively evaluate the performance of the shift scheduling plan in terms of nurse satisfaction, nursing demand coverage rate, nurse fatigue index and the number of shift schedule adjustments. The final fitness synthesizes the values of these two fitness functions, providing a scientific evaluation basis for the optimization of the shift scheduling plan. By maximizing the final fitness, an optimal shift scheduling plan that satisfies both nurse satisfaction and nursing demand coverage rate and minimizes the nurse fatigue index and the number of shift schedule adjustments can be obtained.

[0140] Step 204: Use a genetic algorithm combined with dynamic constraint conditions to generate a shift schedule that meets the hierarchical nurse allocation plan and the nursing demands.

[0141] Among them, the dynamic constraint conditions of the genetic algorithm include daily working hour constraints, weekly shift days constraints and consecutive rest days constraints.

[0142] In the present invention, a genetic algorithm (Genetic Algorithm, GA) is used to generate a shift schedule that meets the hierarchical nurse allocation plan and nursing demands. The genetic algorithm is a search algorithm based on the principles of natural selection and genetics, and solves optimization problems by simulating the biological evolution process.

[0143] Specifically, first, a group of possible shift scheduling plans can be randomly generated, and each plan is called an "individual". Each individual consists of multiple "genes", and each gene represents the shift scheduling information of a nurse in a certain shift.

[0144] The fitness function can be used to evaluate the quality of each individual. The fitness function gives a numerical value based on the quality of the shift scheduling plan (such as nurse satisfaction, nursing demand coverage rate, fatigue index, etc.). The higher the fitness, the better the shift scheduling plan. Then, based on the fitness, some individuals can be selected as the "parents" of the next generation. Usually, methods such as roulette wheel selection and tournament selection are used, and individuals with higher fitness have a higher probability of being selected. Then, some genes of two parent individuals can be exchanged to generate new offspring individuals. For example, the shifts of some nurses in two shift scheduling plans are exchanged to generate a new shift scheduling plan. Finally, the newly generated offspring individuals are randomly mutated, and the value of a certain gene is changed with a certain probability. For example, the shift of a certain nurse is randomly changed to increase the diversity of the population and prevent the algorithm from falling into a local optimum.

[0145] During the process of generating the shift schedule, the following dynamic constraint conditions need to be satisfied:

[0146] Daily working hours constraint: The daily working hours of each nurse do not exceed 8 hours. When generating the shift scheduling plan, ensure that the shift arrangement of each nurse does not exceed the specified daily working hours.

[0147] Weekly shift days constraint: The weekly shift days of each nurse do not exceed 5 days. During the shift scheduling process, count the weekly working days of each nurse to ensure that it does not exceed the specified upper limit.

[0148] Consecutive rest days constraint: Try to ensure that each nurse has at least 2 consecutive rest days per week. When scheduling shifts, give priority to arranging the rest days of nurses to ensure that the requirement of consecutive rest days is met.

[0149] In the process of generating a shift schedule using a genetic algorithm, a group of initial shift schedules that meet the dynamic constraint conditions are randomly generated first. Each schedule is a possible shift schedule, including the shift arrangements of all nurses in different shifts. Then, a fitness function is used to evaluate the quality of each shift schedule. The fitness function comprehensively considers factors such as nurse satisfaction, nursing demand coverage rate, and nurse fatigue index. Then, according to the fitness, some shift schedules are selected as the "parents" of the next generation. The roulette wheel selection method can be used, and the shift schedules with higher fitness have a higher probability of being selected. Then, some genes of two parent shift schedules are exchanged to generate new offspring shift schedules. For example, the shift arrangements of some nurses in two shift schedules are exchanged. Then, random mutations are performed on the newly generated offspring shift schedules, and the shift of a certain nurse is changed with a certain probability. This step helps to increase the diversity of the population and avoid the algorithm falling into a local optimum. Then, it is checked whether the newly generated shift schedule meets the dynamic constraint conditions (daily working hours, number of shift days per week, consecutive rest days). If it does not meet, adjustments are made or it is regenerated. The above steps are repeated until the preset number of iterations is reached or the fitness no longer increases significantly. In each iteration, the population gradually evolves towards a better shift schedule. Finally, the shift schedule with the highest fitness is selected as the final shift schedule. This schedule maximizes nurse satisfaction and nursing demand coverage rate while meeting the dynamic constraint conditions, and minimizes the nurse fatigue index.

[0150] In practical applications, nursing demands may change over time, so the shift schedule needs to be updated regularly. The genetic algorithm can dynamically adjust the shift schedule according to the latest nursing demand prediction (such as the output of an LSTM model) and nurse data (such as fatigue index, work preferences, etc.). For example, if the nursing demand suddenly increases on a certain day, the algorithm can re-optimize the shift schedule to ensure that the demand is met while minimizing the impact on the nurse work arrangements.

[0151] By combining the genetic algorithm with dynamic constraint conditions, the present invention can generate a shift schedule that meets the hierarchical nurse allocation plan and nursing demands. The genetic algorithm gradually optimizes the shift schedule by simulating the natural selection and genetic processes, ensuring that while meeting the constraint conditions, the comprehensive benefits of the shift schedule are maximized. It not only improves the scientificity and rationality of the shift schedule but also enhances the flexibility and adaptability of nursing resources.

[0152] Step 205: Based on the shift data and physiological indicators of the shift schedule, dynamically adjust the workload of each nurse through a supervised learning model to generate a first shift strategy.

[0153] In some embodiments, the physiological indicators include heart rate variability data, the supervised learning model is a random forest classifier, and step 205 may include:

[0154] Calculate the fatigue index of nurses by using the random forest classifier in combination with the shift scheduling data and the heart rate variability data of nurses;

[0155] Dynamically adjust the workload of each nurse according to the fatigue index of the nurse to generate the first shift scheduling strategy.

[0156] In this application, the random forest classifier is used to calculate the fatigue index of nurses by combining the shift scheduling data and the heart rate variability data of nurses, and dynamically adjust the workload of each nurse according to the fatigue index to generate the first shift scheduling strategy.

[0157] The purpose of calculating the fatigue index of nurses by the random forest classifier is to accurately evaluate the fatigue level of nurses by combining the shift scheduling data and the heart rate variability data of nurses through the random forest classifier.

[0158] The input data can be shift scheduling data: including the working hours of nurses, night shift frequency, consecutive working days, rest days, etc. And heart rate variability data: the heart rate variability data of nurses collected through devices such as smart bracelets, reflecting the physiological fatigue state of nurses.

[0159] Random Forest is an ensemble learning algorithm that improves the accuracy of classification or regression by constructing multiple decision trees. In this solution, random forest is used to classify the fatigue state of nurses.

[0160] In the data preparation stage, the shift scheduling data and the heart rate variability data can be integrated into a feature vector. For example, the feature vector may include: working hours, night shift frequency, consecutive working days, heart rate variability metrics (such as SDNN, RMSSD, etc.).

[0161] In the model training stage, historical data (nurse data with known fatigue states) can be used to train the random forest classifier. The training data should contain the above features and corresponding fatigue labels (such as low fatigue, medium fatigue, high fatigue).

[0162] In the fatigue index calculation stage, for each nurse, the trained random forest classifier can be used to predict their fatigue state. According to the prediction results, the fatigue index is calculated. For example, the fatigue state can be mapped to a numerical value: low fatigue: 1, medium fatigue: 2, high fatigue: 3. The output result can be the fatigue index, and the fatigue index of each nurse reflects their current fatigue level.

[0163] Dynamically adjust the workload of each nurse according to the fatigue index. The purpose is to dynamically adjust their workload according to the fatigue index of nurses to ensure the reasonable work arrangement of nurses and avoid over-fatigue.

[0164] Workload assessment evaluates the current workload of each nurse based on the fatigue index. For example, nurses with a high fatigue index should have their workload reduced, while nurses with a low fatigue index can have their workload appropriately increased. The dynamic adjustment strategy adjusts the nurse scheduling plan according to the fatigue index. The adjustment strategy can include: reducing the number of night shifts for highly fatigued nurses, increasing the working hours of low-fatigue nurses, arranging rest days for highly fatigued nurses, and optimizing the shift assignment of nurses to ensure a balanced workload.

[0165] Finally, based on the adjusted nurse workload, the first scheduling strategy can be generated. This strategy should meet the following conditions: comply with dynamic constraints (such as daily working hours, number of days scheduled per week, consecutive rest days), maximize nurse satisfaction, and ensure the coverage of nursing needs.

[0166] The output result can be the first scheduling strategy, which is a detailed schedule that clearly shows the assignment of each nurse to different shifts, taking into account the nurse's fatigue index to ensure a reasonable workload.

[0167] An example is given. Suppose there are the following data and models:

[0168] The input data is the scheduling data:

[0169] Nurse A: Working hours = 8 hours, night shift frequency = 2 times / week, consecutive working days = 5 days;

[0170] Nurse B: Working hours = 7 hours, night shift frequency = 1 time / week, consecutive working days = 4 days.

[0171] Heart rate variability data:

[0172] Nurse A: Heart rate variability index SDNN = 50 ms, RMSSD = 30 ms;

[0173] Nurse B: Heart rate variability index SDNN = 60 ms, RMSSD = 40 ms.

[0174] Random forest classifier, the trained random forest model predicts the fatigue status based on the input features:

[0175] Nurse A: Fatigue status = high fatigue (fatigue index = 3);

[0176] Nurse B: Fatigue status = low fatigue (fatigue index = 1).

[0177] Dynamic adjustment strategy, adjusting the nurse's workload according to the fatigue index:

[0178] Nurse A: Reduce the number of night shifts and arrange 1 day of rest;

[0179] Nurse B: Appropriately increase the working hours and arrange 1 night shift.

[0180] The shift schedule generated by the first shift scheduling strategy is as follows:

[0181] Nurse A: Working hours = 7 hours, night shift frequency = 1 time / week, consecutive working days = 4 days, rest days = 2 days;

[0182] Nurse B: Working hours = 8 hours, night shift frequency = 2 times / week, consecutive working days = 5 days, rest days = 2 days.

[0183] In some embodiments, the method of the present invention can also automatically identify legal holidays, nurses' birthdays, and nurses' reserved vacation times. Specifically, by automatically identifying legal holidays, nurses' birthdays, and nurses' reserved vacation times, the shift scheduling plan can be further optimized to ensure that while meeting the nursing needs, the personal time of nurses is fully respected, and the job satisfaction and quality of life of nurses are improved.

[0184] In specific implementation, the national legal holiday information, including holiday dates and types (such as Spring Festival, National Day, etc.), can be obtained from the hospital information system or external calendar services. The birthday information of nurses can be extracted from the hospital's human resources management system to ensure that the system can identify the birthday dates of each nurse. Nurses are allowed to submit reserved vacation applications through the hospital's internal system, and the system automatically records and manages these reserved vacation times.

[0185] In some embodiments, legal holidays, nurses' birthdays, and reserved vacation times can be integrated into the data model of the shift scheduling system to ensure that these special dates can be considered in real time when generating the shift schedule. For legal holidays, the system automatically marks these dates and makes special arrangements according to the hospital's shift scheduling policies (such as work arrangements on legal holidays, overtime compensation, etc.). For nurses' birthdays, the system gives priority to arranging rest or relatively easy shifts for nurses during scheduling, reflecting the care for nurses. For reserved vacation times, the system automatically avoids these dates when generating the shift schedule to ensure that nurses can take vacations according to the reservation.

[0186] When generating the shift schedule, the system can automatically check the legal holidays, birthdays, and reserved vacation times of each nurse to ensure that the shift scheduling plan meets these constraints. If a certain nurse has a scheduling requirement on a legal holiday or birthday, the system will give priority to arranging other nurses to substitute, or adjust the scheduling plan to ensure that nurses can get appropriate rest on these special dates. For reserved vacation times, the system will adjust the scheduling in advance to ensure that during the nurses' vacations, the nursing work can proceed smoothly and avoid chaos caused by temporary scheduling adjustments.

[0187] By automatically identifying and processing these special dates, the system not only improves the scientificity and rationality of shift scheduling, but also demonstrates respect and care for nurses' personal time, contributing to increased job satisfaction and loyalty among nurses. The system can generate reports regularly to remind managers to pay attention to nurses' legal holidays, birthdays, and scheduled vacations, further optimizing human resource management.

[0188] In summary, through the random forest classifier combined with shift scheduling data and heart rate variability data, the present invention can accurately evaluate the fatigue index of nurses, dynamically adjust the workload of each nurse according to the fatigue index, and generate a reasonable first shift scheduling strategy. It not only takes into account the physiological and psychological states of nurses, but also ensures the efficiency and quality of nursing work, improving nurses' job satisfaction and patient care experience.

[0189] Step 206: Automatically match substitute nurses in case of emergencies through the reinforcement learning model and optimize the first shift scheduling strategy to generate a second shift scheduling strategy.

[0190] In some embodiments, the reinforcement learning model is a Deep Q-Learning model, and step 206 may include:

[0191] Obtain the linear combination of the skill matching degree, fatigue index, and shift scheduling time conflict rate of each substitute nurse;

[0192] Optimize the first shift scheduling strategy according to the linear combination of the skill matching degree, fatigue index, and shift scheduling time conflict rate of each substitute nurse, with the optimization goal of minimizing the decline in nursing quality and the additional workload of nurses in the current shift scheduling, and generate the second shift scheduling strategy.

[0193] Specifically, in order to optimize the first shift scheduling strategy and generate a second shift scheduling strategy, it is necessary to comprehensively consider the skill matching degree, fatigue index, and shift scheduling time conflict rate of substitute nurses.

[0194] The purpose of obtaining the linear combination is to evaluate the comprehensive adaptability of each nurse as a substitute by calculating the linear combination of the skill matching degree, fatigue index, and shift scheduling time conflict rate of each substitute nurse.

[0195] The input data can be the skill matching degree: measuring the matching degree between the skills of the substitute nurse and the current nursing needs. For example, an orthopedic nurse has a high skill matching degree for the care of patients after fracture surgery. And the fatigue index: calculated through the random forest classifier combined with shift scheduling data and heart rate variability data, reflecting the current fatigue level of the nurse. And the shift scheduling time conflict rate: measuring the conflict degree between the shift scheduling time of the substitute nurse and the current shift scheduling requirements. For example, if the substitute nurse already has a scheduled shift, the conflict rate will increase.

[0196] First, the weights of skill matching degree, fatigue index, and scheduling time conflict rate can be set. For each substitute nurse, the linear combination of their skill matching degree, fatigue index, and scheduling time conflict rate can be calculated. The output result can be the comprehensive adaptability score, and the comprehensive adaptability score of each substitute nurse reflects their comprehensive adaptability as a substitute.

[0197] The purpose of optimizing the first scheduling strategy is to optimize the first scheduling strategy according to the comprehensive adaptability score, minimizing the decline in nursing quality and the additional workload of nurses in the current scheduling. The optimization goals can include minimizing the decline in nursing quality: ensuring a high match between the skills of substitute nurses and nursing needs, and reducing the decline in nursing quality caused by skill mismatch. And minimizing the additional workload of nurses: avoiding arranging nurses who are overly fatigued or have existing scheduling conflicts, and reducing the additional workload of nurses.

[0198] First, the substitute nurses can be sorted according to the comprehensive adaptability score, and nurses with a high comprehensive adaptability score can be preferentially selected as substitutes. For each shift that needs to be substituted, select the nurse with the highest comprehensive adaptability as a substitute, while ensuring that the scheduling times do not conflict. If the comprehensive adaptability scores of multiple nurses are similar, other factors (such as nurses' work preferences, consecutive working days, etc.) can be further considered for fine-tuning. Then, the adjusted scheduling plan can be evaluated to ensure that dynamic constraints (such as daily working hours, weekly scheduling days, consecutive rest days) are met. If the substitution selection for a certain shift leads to a decline in nursing quality or an increase in the additional workload of nurses, re-adjust the selection to find a sub-optimal solution. The output result can be the second scheduling strategy, that is, the optimized schedule, which clarifies the allocation of each nurse in different shifts, while minimizing the decline in nursing quality and the additional workload of nurses.

[0199] An example is illustrated. Suppose there are the following data and models:

[0200] The input data is the information of substitute nurses:

[0201] Nurse A: Skill matching degree = 0.9, Fatigue index = 2, Scheduling time conflict rate = 0.1;

[0202] Nurse B: Skill matching degree = 0.7, Fatigue index = 1, Scheduling time conflict rate = 0.2;

[0203] Nurse C: Skill matching degree = 0.8, Fatigue index = 3, Scheduling time conflict rate = 0.1.

[0204] Weight setting:

[0205] w1 = 0.5, w2 = 0.3, w3 = 0.2.

[0206] Linear combination calculation:

[0207] Comprehensive adaptability of Nurse A:

[0208] 0.5×0.9−0.3×2−0.2×0.1 = 0.45−0.6−0.02 = −0.17;

[0209] Comprehensive adaptability of Nurse B:

[0210] 0.5×0.7−0.3×1−0.2×0.2 = 0.35−0.3−0.04 = 0.01;

[0211] Comprehensive adaptability of Nurse C:

[0212] 0.5×0.8−0.3×3−0.2×0.1 = 0.4−0.9−0.02 = −0.52.

[0213] Optimized scheduling result:

[0214] Sorted according to the comprehensive adaptability score:

[0215] Nurse B (0.01);

[0216] Nurse A (−0.17);

[0217] Nurse C (−0.52).

[0218] Select Nurse B as the substitute, whose comprehensive adaptability score is the highest.

[0219] Second scheduling strategy:

[0220] The generated schedule is as follows:

[0221] Nurse A: Keep the original schedule;

[0222] Nurse B: Substitute into the shift that needs to be adjusted;

[0223] Nurse C: Keep the original schedule.

[0224] By calculating the linear combination of the skill matching degree, fatigue index and scheduling time conflict rate of each substitute nurse, the present invention can comprehensively evaluate the adaptability of each nurse as a substitute. Optimize the first scheduling strategy according to the comprehensive adaptability score to generate the second scheduling strategy, ensuring that while meeting the nursing needs, the decline in nursing quality and the additional workload of nurses are minimized. It not only improves the scientificity and rationality of scheduling, but also enhances the flexibility and adaptability of nursing resources.

[0225] Please refer to Figure 3 , Figure 3 which is a schematic structural diagram of a clinical intelligent scheduling system based on big data analysis provided by the present invention.

[0226] AsFigure 3 As shown in the figure, a clinical intelligent scheduling system based on big data analysis proposed in an embodiment of the present invention includes:

[0227] A data acquisition module 301, configured to acquire a data set of patient data, nurse data, and environmental data, and preprocess the data set to generate a standardized data set;

[0228] A demand prediction module 302, configured to process the standardized data set by using a time series prediction model and a decision tree model to predict the nursing demand for future scheduling;

[0229] A hierarchical allocation module 303, configured to generate a hierarchical nurse allocation plan based on the skills of nurses, the current workload, and historical work data through a recommendation algorithm;

[0230] A preliminary scheduling module 304, configured to generate a schedule that meets the hierarchical nurse allocation plan and the nursing demand by using a genetic algorithm combined with dynamic constraints;

[0231] A first scheduling module 305, configured to dynamically adjust the workload of each nurse based on the scheduling data and physiological indicators of the schedule through a supervised learning model to generate a first scheduling strategy;

[0232] A second scheduling module 306, configured to automatically match substitute nurses and optimize the first scheduling strategy in case of emergencies through a reinforcement learning model to generate a second scheduling strategy.

[0233] Please refer to Figure 4 , Figure 4 which is a schematic diagram of an embodiment of an electronic device provided by an embodiment of the present invention. As Figure 4 shown, an embodiment of the present invention provides an electronic device 400, including a memory 410, a processor 420, and a computer program 411 stored on the memory 410 and executable on the processor 420. When the processor 420 executes the computer program 411, the following steps are implemented:

[0234] Acquire a data set of patient data, nurse data, and environmental data, and preprocess the data set to generate a standardized data set;

[0235] Process the standardized data set by using a time series prediction model and a decision tree model to predict the nursing demand for future scheduling;

[0236] Generate a hierarchical nurse allocation plan based on the skills of nurses, the current workload, and historical work data through a recommendation algorithm;

[0237] Generate a schedule that meets the hierarchical nurse allocation plan and the nursing demand by using a genetic algorithm combined with dynamic constraints;

[0238] Based on the scheduling data and physiological indicators of the said scheduling table, dynamically adjust the workload of each nurse through a supervised learning model to generate a first scheduling strategy;

[0239] Automatically match substitute nurses in case of emergencies through a reinforcement learning model and optimize the said first scheduling strategy to generate a second scheduling strategy.

[0240] Please refer to Figure 5 , Figure 5 which is a schematic diagram of an embodiment of a computer-readable storage medium provided by an embodiment of the present invention. As Figure 5 shown, this embodiment provides a computer-readable storage medium 500, on which a computer program 411 is stored. When the computer program 411 is executed by a processor, the following steps are implemented:

[0241] Obtain a data set of patient data, nurse data, and environmental data, and preprocess the said data set to generate a standardized data set;

[0242] Use a time series prediction model and a decision tree model to process the said standardized data set to predict the nursing needs for future scheduling;

[0243] Based on the skills, current workload, and historical work data of nurses, generate a hierarchical nurse allocation plan through a recommendation algorithm;

[0244] Adopt a genetic algorithm combined with dynamic constraint conditions to generate a scheduling table that meets the said hierarchical nurse allocation plan and the said nursing needs;

[0245] Based on the scheduling data and physiological indicators of the said scheduling table, dynamically adjust the workload of each nurse through a supervised learning model to generate a first scheduling strategy;

[0246] Automatically match substitute nurses in case of emergencies through a reinforcement learning model and optimize the said first scheduling strategy to generate a second scheduling strategy.

[0247] It should be noted that in the above embodiments, the descriptions of each embodiment have their own emphases. For the parts not detailed in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0248] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0249] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present invention. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and combinations of flows and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to the processors of general-purpose computers, special-purpose computers, embedded computers, or other programmable data processing devices to produce a machine, such that the instructions executed by the processors of the computer or other programmable data processing devices produce a system for implementing the functions specified in one or more flows Figure 1 one or more flows and / or blocks Figure 1 or more blocks.

[0250] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory produce a manufactured article including an instruction system that implements the functions specified in one or more flows Figure 1 one or more flows and / or blocks Figure 1 or more blocks.

[0251] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one or more flows Figure 1 one or more flows and / or blocks Figure 1 or more blocks.

[0252] Although the preferred embodiments of the present invention have been described, those skilled in the art can make additional changes and modifications once they know the basic inventive concept. Therefore, the appended claims are intended to be construed to include the preferred embodiments and all changes and modifications falling within the scope of the present invention.

[0253] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these changes and modifications.

Claims

1. A clinical intelligent scheduling method based on big data analysis, characterized in that, The method includes: Obtaining a data set of patient data, nurse data, and environmental data, and preprocessing the data set to generate a standardized data set; Using a time series prediction model and a decision tree model to process the standardized data set to predict the nursing needs for future scheduling; Based on the skills of nurses, current workload, and historical work data, generating a hierarchical nurse allocation plan through a recommendation algorithm, including: a collaborative filtering recommendation algorithm to match the fitness of each nurse to the scheduling shift based on historical scheduling data; a KNN algorithm to calculate the workload similarity between nurses; generating the hierarchical nurse allocation plan according to the fitness and the workload similarity; also including: constructing a first fitness function for maximizing the weighted sum of the nurse satisfaction score and the nursing needs coverage rate; constructing a second fitness function for minimizing the product of the total nurse fatigue index and the number of scheduling adjustments; determining the fitness according to the function value of the first fitness function and the function value of the second fitness function; Using a genetic algorithm combined with dynamic constraint conditions to generate a schedule that meets the hierarchical nurse allocation plan and the nursing needs; the dynamic constraint conditions of the genetic algorithm include daily working hours constraint, weekly scheduling days constraint, and consecutive rest days constraint; Based on the scheduling data and physiological indicators of the schedule, dynamically adjusting the workload of each nurse through a supervised learning model to generate a first scheduling strategy; Automatically matching substitute nurses and optimizing the first scheduling strategy in case of emergencies through a reinforcement learning model to generate a second scheduling strategy.

2. The clinical intelligent scheduling method based on big data analysis according to claim 1, wherein The preprocessing of the data set to generate a standardized data set includes: Using the K-means clustering algorithm to process the nurse data, grouping according to the skill level and work intensity of nurses to obtain the grouped nurse data; Using the PCA dimensionality reduction algorithm to reduce the dimensionality of the multi-dimensional feature data of the grouped nurse data, the patient data, and the environmental data and extract key features to obtain the standardized data set.

3. The clinical intelligent scheduling method based on big data analysis according to claim 2, wherein, The time series prediction model is an LSTM network, and the decision tree model is an XGBoost algorithm; The using of the time series prediction model and the decision tree model to process the standardized data set to predict the nursing needs for future scheduling includes: Using the LSTM network to process the standardized data set to predict the nursing needs fluctuation of postoperative patients at night; Using the XGBoost algorithm to process the standardized data set to predict the nursing needs level for future scheduling; Determining the nursing needs for future scheduling according to the nursing needs fluctuation and the nursing needs level.

4. The clinical intelligent scheduling method based on big data analysis according to claim 1, characterized in that The physiological indicators include heart rate variability data, and the supervised learning model is a random forest classifier; The based on the scheduling data and physiological indicators of the schedule, dynamically adjusting the workload of each nurse through a supervised learning model to generate a first scheduling strategy includes: Using the random forest classifier combined with the scheduling data and the heart rate variability data of nurses to calculate the fatigue index of nurses; Dynamically adjust the workload of each nurse according to the fatigue index of the nurse to generate the first shift scheduling strategy.

5. The clinical intelligent scheduling method based on big data analysis according to claim 4, characterized in that, The reinforcement learning model is a Deep Q-Learning model; automatically matching substitute nurses and optimizing the first shift scheduling strategy through the reinforcement learning model in case of emergencies to generate the second shift scheduling strategy, including: Obtain the linear combination of the skill matching degree, fatigue index and shift scheduling time conflict rate of each substitute nurse. Optimize the first shift scheduling strategy according to the linear combination, and the optimization goal is to minimize the decrease in nursing quality and the additional workload of nurses in the current shift scheduling to generate the second shift scheduling strategy.

6. The clinical intelligent scheduling method based on big data analysis according to claim 5, wherein The nurse data includes the skill level, shift scheduling preference, night shift acceptance flag and historical work log of the nurse, and the environmental data includes the predicted increase in nursing demand during the peak season of seasonal diseases.

7. A clinical intelligent scheduling system based on big data analysis, characterized in that, The system includes: A data acquisition module, configured to acquire a data set of patient data, nurse data and environmental data, and preprocess the data set to generate a standardized data set. A demand prediction module, configured to use a time series prediction model and a decision tree model to process the standardized data set to predict the nursing demand for future shift scheduling. A hierarchical allocation module, configured to generate a hierarchical nurse allocation plan through a recommendation algorithm based on the skills, current workload and historical work data of the nurses, and is also used for a collaborative filtering recommendation algorithm to match the fitness of each nurse and the shift schedule based on historical shift scheduling data; a KNN algorithm to calculate the workload similarity between nurses; generate the hierarchical nurse allocation plan according to the fitness and the workload similarity; is also used to construct a first fitness function for maximizing the weighted sum of the nurse satisfaction score and the nursing demand coverage rate; construct a second fitness function for minimizing the product of the total nurse fatigue index and the number of shift schedule adjustments; determine the fitness according to the function value of the first fitness function and the function value of the second fitness function. A preliminary shift scheduling module, configured to generate a shift schedule that meets the hierarchical nurse allocation plan and the nursing demand by using a genetic algorithm combined with dynamic constraints; the dynamic constraints of the genetic algorithm include daily working hour constraints, weekly shift scheduling days constraints and consecutive rest days constraints. A first shift scheduling module, configured to dynamically adjust the workload of each nurse based on the shift scheduling data and physiological indicators of the shift schedule to generate a first shift scheduling strategy. A second shift scheduling module, configured to automatically match substitute nurses and optimize the first shift scheduling strategy through a reinforcement learning model in case of emergencies to generate a second shift scheduling strategy.

Images