A method, system, device, and medium for joint optimization of vital sign monitoring tasks

Abstract

The application relates to a sign monitoring task joint optimization method, system, device and medium. In order to solve the problem of achieving the optimal balance between the diversified needs of the upper clinical monitoring task and the resource constraints of the underlying sensor, the aggregation process of different types of physiological data from the sensor to the terminal is abstracted as a multi-commodity network flow problem. A target function is built based on the difference between the minimum sensor energy consumption and the additional clinical value of the secondary monitoring task. Corresponding constraint conditions and decision variable items are set. The monitoring task decision and sensor resource allocation are combined to build a joint optimization problem. By solving the joint optimization problem, the monitoring task decision and sensor data resource allocation joint optimization can be realized, the effective execution of the core monitoring task is ensured, and the operation utility and life cycle are maximized.

Description

Technical Field

[0001] This application relates to the technical field, and in particular to a joint optimization method, system, device and medium for vital sign monitoring tasks. Background Technology

[0002] With the development of wearable devices and telemedicine technologies, human body monitoring systems continuously collect physiological data through multiple sensors deployed in various parts of the body, providing crucial support for health monitoring and disease early warning. These platforms typically include various physiological signal acquisition sensors (such as ECG, body temperature, and blood oxygen sensors) and a central processing terminal (such as a gateway). The sensors are usually responsible for collecting raw physiological data and transmitting it to the terminal for further analysis and monitoring tasks.

[0003] However, existing technologies often separate task scheduling from data transmission, resulting in upper-level monitoring task decisions being unaware of the energy consumption costs of lower-level sensors, or lower-level sensors failing to understand the value of upper-level monitoring tasks in their data transmission, leading to resource waste or a decline in monitoring quality. Summary of the Invention

[0004] The following is an overview of the subject matter described in detail herein. This overview is not intended to limit the scope of the claims.

[0005] The main objective of this disclosure is to propose a joint optimization method, system, device, and medium for vital sign monitoring tasks, which can achieve joint optimization of monitoring task decision-making and sensor data resource allocation, ensuring the effective execution of core monitoring tasks while maximizing operational efficiency and lifecycle.

[0006] A first aspect of this application provides a joint optimization method for vital sign monitoring tasks, applied to a human monitoring platform with multiple monitoring tasks, the human monitoring platform comprising:

[0007] Multiple sensors can collect corresponding raw physiological data from the human body, simplify the raw physiological data locally to obtain simplified first physiological data, and transmit the first physiological data to the terminal.

[0008] The terminal is connected to multiple sensors via corresponding links. The terminal can execute at least one of the multiple monitoring tasks, and a monitoring task needs to be executed collaboratively using multiple first physiological data.

[0009] The joint optimization method for the vital signs monitoring task includes:

[0010] The multiple monitoring tasks are divided into core monitoring tasks and secondary monitoring tasks;

[0011] The overall constraints and objective function of the joint optimization problem are determined. The overall objective function is to minimize the difference between a first function and a second function. The first function includes the idle operating energy consumption of the sensor, the computational energy consumption of the sensor performing local simplification, and the transmission energy consumption of the sensor transmitting the corresponding first physiological data to the terminal. The transmission energy consumption is related to the data transmission rate allocated by the sensor on the corresponding link for transmitting the corresponding first physiological data. The second function includes the additional clinical value benefit of performing secondary monitoring tasks. The overall constraints include:

[0012] Condition 1: The basis for the terminal to execute monitoring tasks is that the data flow of each type of first physiological data required for the monitoring task received by the terminal is not less than the preset minimum requirement; Condition 2: All core monitoring tasks must be executed; Condition 3: The total computing resources of the terminal to execute at least one monitoring task do not exceed the terminal's maximum computing resources; Condition 4: The total energy consumption of each sensor does not exceed the sensor's energy consumption limit; where the total energy consumption of the sensor is equal to the sum of the corresponding idle operation energy consumption, computing energy consumption, and transmission energy consumption; Condition 5: The first physiological data after local simplification by the sensor does not exceed the corresponding initial physiological data; Condition 6: The data flow transmitted by the sensor on the corresponding link and the data flow received by the terminal on the corresponding link are conserved; Condition 7: The data flow transmitted in any link does not exceed the maximum capacity of the corresponding link;

[0013] The decision variables of the joint optimization problem are determined, and the joint optimization problem is solved based on the overall constraints and the overall objective function to obtain the optimal solution of the decision variables. The decision variables of the joint optimization problem include: a binary decision variable, the data transmission rate allocated by the sensor in the corresponding link for transmitting the corresponding first physiological data, and the amount of data of the corresponding initial physiological data that is locally simplified by the sensor. The binary decision variable uses 0 or 1 to represent whether each of the multiple monitoring tasks is executed.

[0014] This embodiment provides a joint optimization method for a vital signs monitoring task, which has at least the following beneficial effects:

[0015] To address the challenge of achieving the optimal balance between the diverse needs of upper-level clinical monitoring tasks and the resource constraints of lower-level sensors, this method abstracts the convergence process of different types of physiological data from sensors to terminals as a multi-commodity network flow problem. In scenarios involving multiple monitoring tasks, an objective function is established based on minimizing the difference between sensor energy consumption and the additional clinical value gains of secondary monitoring tasks. Corresponding overall constraints and decision variables are set, and a joint optimization problem is constructed by combining monitoring task decision-making with sensor resource allocation. By solving the joint optimization problem, the joint optimization of monitoring task decision-making and sensor data resource allocation can be achieved, ensuring the effective execution of core monitoring tasks while maximizing operational utility and lifecycle.

[0016] A second aspect of this application provides a joint optimization system for vital sign monitoring tasks, the system comprising:

[0017] A human body monitoring platform; the human body monitoring platform includes multiple sensors and terminals. The sensors can collect corresponding raw physiological data from the human body, simplify the raw physiological data locally to obtain simplified first physiological data, and transmit the first physiological data to the terminal. The terminal is connected to the multiple sensors through corresponding links. The terminal can execute at least one of multiple monitoring tasks, and a monitoring task needs to be executed collaboratively using multiple first physiological data.

[0018] A server, the server comprising:

[0019] The problem-setting module is used to classify the multiple monitoring tasks into core monitoring tasks and secondary monitoring tasks;

[0020] The objective function determination module is used to determine the overall constraints and overall objective function of the joint optimization problem. The overall objective function is to minimize the difference between a first function and a second function. The first function includes the sensor's idle operating energy consumption, the sensor's computational energy consumption for local simplification, and the sensor's transmission energy consumption for transmitting the corresponding first physiological data to the terminal. The transmission energy consumption is related to the data transmission rate allocated by the sensor on the corresponding link for transmitting the corresponding first physiological data. The second function includes the additional clinical value benefit of performing secondary monitoring tasks. The overall constraints include:

[0021] Condition 1: The basis for the terminal to execute monitoring tasks is that the data flow of each type of first physiological data required for the monitoring task received by the terminal is not less than the preset minimum requirement; Condition 2: All core monitoring tasks must be executed; Condition 3: The total computing resources of the terminal to execute at least one monitoring task do not exceed the terminal's maximum computing resources; Condition 4: The total energy consumption of each sensor does not exceed the sensor's energy consumption limit; where the total energy consumption of the sensor is equal to the sum of the corresponding idle operation energy consumption, computing energy consumption, and transmission energy consumption; Condition 5: The first physiological data after local simplification by the sensor does not exceed the corresponding initial physiological data; Condition 6: The data flow transmitted by the sensor on the corresponding link and the data flow received by the terminal on the corresponding link are conserved; Condition 7: The data flow transmitted in any link does not exceed the maximum capacity of the corresponding link;

[0022] The problem-solving module is used to determine the decision variables of the joint optimization problem and solve the joint optimization problem based on the overall constraints and the overall objective function to obtain the optimal solution of the decision variables. The decision variables of the joint optimization problem include: a binary decision variable, the data transmission rate allocated by the sensor in the corresponding link for transmitting the corresponding first physiological data, and the amount of data of the corresponding initial physiological data that is locally simplified by the sensor. The binary decision variable uses 0 or 1 to represent whether each of the multiple monitoring tasks is executed.

[0023] A third aspect of this application provides an electronic device including at least one controller and a memory for communicatively connecting to the controller; the memory stores instructions executable by the at least one controller to cause the at least one controller to perform a joint optimization method for a vital signs monitoring task as described above.

[0024] A fourth aspect of this application provides a computer-readable storage medium storing computer-executable instructions for causing a computer to perform a joint optimization method for a vital signs monitoring task as described above.

[0025] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0026] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0027] Figure 1 This is a flowchart illustrating an embodiment of a joint optimization method for a vital signs monitoring task provided in this application;

[0028] Figure 2 This is a schematic diagram of the structure of an embodiment of a joint optimization system for vital sign monitoring tasks provided in this application;

[0029] Figure 3 This is a schematic diagram of the structure of an embodiment of an electronic device provided in this application. Detailed Implementation

[0030] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0031] One embodiment of this application provides a joint optimization method for vital sign monitoring tasks, applied to a human monitoring platform with multiple monitoring tasks. The human monitoring platform includes:

[0032] Multiple sensors can collect corresponding raw physiological data from the human body, simplify the raw physiological data locally to obtain simplified first physiological data, and transmit the first physiological data to the terminal.

[0033] The terminal is connected to multiple sensors via corresponding links. The terminal can execute at least one of multiple monitoring tasks, and a monitoring task needs to be executed collaboratively using multiple primary physiological data.

[0034] The joint optimization method for the vital sign monitoring task in this embodiment includes:

[0035] Step S100: Divide the multiple monitoring tasks into core monitoring tasks and secondary monitoring tasks;

[0036] Step S200: Determine the overall constraints and overall objective function of the joint optimization problem; the overall objective function is to minimize the difference between the first function and the second function. The first function includes the idle operating energy consumption of the sensor, the computational energy consumption of the sensor performing local simplification, and the transmission energy consumption of the sensor transmitting the corresponding first physiological data to the terminal; the transmission energy consumption is related to the data transmission rate allocated by the sensor on the corresponding link for transmitting the corresponding first physiological data; the second function includes the additional clinical value benefit of performing secondary monitoring tasks; the overall constraints include:

[0037] Condition 1: The basis for the terminal to execute monitoring tasks is that the data flow of each type of first physiological data required for the monitoring task received by the terminal is not less than the preset minimum requirement; Condition 2: All core monitoring tasks must be executed; Condition 3: The total computing resources of the terminal to execute at least one monitoring task do not exceed the terminal's maximum computing resources; Condition 4: The total energy consumption of each sensor does not exceed the sensor's energy consumption limit; where the total energy consumption of the sensor is equal to the sum of the corresponding idle operation energy consumption, computing energy consumption, and transmission energy consumption; Condition 5: The first physiological data after local simplification by the sensor does not exceed the corresponding initial physiological data; Condition 6: The data flow transmitted by the sensor on the corresponding link and the data flow received by the terminal on the corresponding link are conserved; Condition 7: The data flow transmitted in any link does not exceed the maximum capacity of the corresponding link;

[0038] Step S300: Determine the decision variables of the joint optimization problem, and solve the joint optimization problem based on the overall constraints and the overall objective function to obtain the optimal solution of the decision variables. The decision variables of the joint optimization problem include: binary decision variables, the data transmission rate allocated by the sensor in the corresponding link for transmitting the corresponding first physiological data, and the amount of data of the corresponding initial physiological data that is locally simplified by the sensor. The binary decision variables use 0 or 1 to represent whether each monitoring task in the multiple monitoring tasks is executed.

[0039] First, let's introduce the application scenario of this embodiment:

[0040] This embodiment focuses on using a Wireless Body Area Network (WBAN) to realize home monitoring of patients with chronic diseases. When multiple monitoring tasks exist, the human body monitoring platform rationally selects at least one monitoring task to be executed. These monitoring tasks include core monitoring and secondary monitoring tasks. The human body monitoring platform includes at least:

[0041] Deployed in various parts of the patient's body A set of sensors ); each sensor It can only generate one or more specific types of physiological data that it pre-defines. (Using a set) To represent the sensor It can generate a set of physiological data types, such as heart rate and body temperature. A terminal that acts as a personal health gateway. ;

[0042] The execution entity in this embodiment is a server. Servers and terminals in the platform Communication connection.

[0043] The human body monitoring platform performs monitoring tasks in two main phases. The first phase is the data aggregation phase, in which various sensors... Generate different types of physiological data ( These physiological data streams converge at the terminal via corresponding links. The second stage is the task processing stage, on the terminal. By leveraging a variety of aggregated physiological data, multiple monitoring tasks can be selected and executed (ensemble) Examples include "arrhythmia detection," which integrates and refines collected physiological data into clinically valuable diagnostic information.

[0044] It should be noted that any monitoring task Its execution depends on a specific set of physiological data (set) required by it. The main objective of this application is to seek the optimal balance between system resources (sensor power consumption) and clinical monitoring effectiveness. It is important to note that this does not mean that multiple monitoring tasks are executed simultaneously, but rather that appropriate secondary monitoring tasks are selected and executed only when the core monitoring task must be performed, in order to achieve the optimal balance.

[0045] The following section introduces the construction of the problem model, as well as the determination of the objective function, constraints, and decision variables;

[0046] First of all, let The first physiological data received by the terminal from the corresponding link Total aggregate traffic, The calculation method will be explained later. It should be noted that the physiological data collected by the sensor is called raw physiological data, and the physiological data after simplification (such as feature extraction, noise reduction, deduplication and redundancy removal, compression, etc.) is called first physiological data. The purpose of simplification is to reduce the amount of data transmitted, thereby reducing energy consumption.

[0047] Then, binary decision variables are introduced. To indicate the monitoring task Whether it was successfully executed, if Then the task It will be executed, otherwise This indicates that the task has not been executed. Whether a task is executed depends on the terminal. Does the received flow rate of all required physiological data meet the preset minimum requirements (i.e., minimum clinical analysis requirements)? This logical relationship constitutes constraint one (C1) of the problem model:

[0048] For any monitoring task and every type of first physiological data required Its successful execution ( The basis for this is that the corresponding data traffic must meet the preset minimum requirements, which is constrained by the following formula:

[0049] ;

[0050] Condition 2 (C2): All core monitoring tasks It must be executed, that is:

[0051] ;

[0052] In addition, considering the terminal Due to computational resource limitations, let's assume that each monitoring task is executed... Requires to occupy The computing resources, and the terminal The maximum computing resources available for a terminal used for monitoring tasks are Therefore, the total computing resources required for a terminal to execute at least one monitoring task (i.e., to decide which monitoring task to execute) do not exceed the terminal's maximum computing resources.

[0053] Condition 3 (C3): Gateway computing capacity constraint:

[0054] ;

[0055] For any sensor Its instantaneous power consumption consists of three parts:

[0056] 1) Maintain idle operating power 2) The computational power for simplifying the raw physiological data (such as compression or feature extraction); where the computational power is proportional to the amount of raw physiological data being processed; 3) The data obtained after simplifying the raw physiological data locally is called the first physiological data, and the first physiological data is transmitted to the terminal via the corresponding link. The transmission power depends on the actual data transmission rate allocated by the sensor on the corresponding link for transmitting the corresponding first physiological data. , and For link The two endpoints on the network. It is important to note that the WBAN network only involves the sensor transmitting the first physiological data to the terminal; it does not involve data transmission between sensors, nor does it involve the terminal transmitting physiological data to the sensor.

[0057] In summary, the sensor During the work cycle Total energy consumption within for:

[0058] ;

[0059] here, Defined as a sensor The calculation of the energy consumption coefficient, i.e., the sensor The energy consumption coefficient required to simplify the raw physiological data. It is a comprehensive parameter that represents a specific period. Inside, from the sensor The average or conservative estimate of the energy required to transmit a unit of data; it abstracts the physical layer details such as the underlying channel state and transmission distance. The amount of raw physiological data, or the amount of data from sensors. Generate the first physiological data The rate.

[0060] Given the physical limitations of sensor battery capacity, the total energy consumption of each sensor must be within its energy consumption limit. Within the scope, that is:

[0061] Condition 4 (C4): The total power consumption of each sensor must not exceed :

[0062] ;

[0063] In C1 above The calculation method is as follows: all inflow terminals The link is used to transmit the first physiological data. The sum of the allocated data transfer rates:

[0064] ;

[0065] For any directed link , This indicates that the data was assigned to the first physiological data. The average transmission rate (units can be bit / s or sample / s). Among them, This refers to the node Flow to terminal First physiological data Data traffic is special case.

[0066] This embodiment abstracts the aggregation process of different types of primary physiological data from sensors to terminals as a multi-goods network flow problem. In this problem model, different types of primary physiological data, such as ECG and SpO2, are considered different goods. This is because although they share the common physical transport network WBAN (i.e., share link bandwidth), their information content is different and they cannot be substituted for each other; therefore, they need to be tracked and aggregated independently in the analysis task. In this network… In, each sensor It is the specific first physiological data it produces. The source point, and the terminal It is the common convergence point of all these primary physiological data. Sensors Generate raw physiological data The original rate is Before the first physiological data enters the link for transmission, the sensor The amount of data that needs to be transmitted can be reduced through local decompression (such as data compression or digest extraction). Here, variables are used. This refers to the amount of raw physiological data that has been simplified locally, or the data rate.

[0067] Condition 5 (C5): The amount of data simplified locally on the sensor cannot exceed the amount of its original physiological data.

[0068] ;

[0069] Therefore, sensor For the first physiological data Net supply of sensors The data rate of the injected link is .

[0070] Every type of primary physiological data flowing in the link All must satisfy the law of flow conservation, which states that for any node (including sensors or terminals), the total inflow of data equals the total outflow of data plus the node's net supply (or net consumption). This constitutes the core constraint:

[0071] Condition 6 (C6): For any node and any first physiological data Its flow must be conserved:

[0072] ;

[0073] in, Represents a node Total inflow, Represents a node Total outflow. C6 is used to characterize the inflow to each node and each first physiological data point = outflow + net supply.

[0074] Due to limited channel bandwidth, the total traffic of all first physiological data carried on any single link cannot exceed the maximum capacity of that link. :

[0075] Condition 7 (C7): In this model, It can be set based on average channel quality or as a conservative bandwidth to ensure quality of service (QoS).

[0076] ;

[0077] In summary, the goal is to find an optimal combination of decision variables. ,in , , The decision combination must satisfy all the constraints C1 to C6 set above, as well as the range constraints of the variables themselves:

[0078] ;

[0079] The core objective of the model is to maximize overall efficiency while ensuring that core monitoring tasks are executed and sensor energy consumption does not exceed the budget. This manifests as minimizing the total operating energy consumption of the system, while incentivizing the execution of as many valuable non-core monitoring tasks (i.e., secondary monitoring tasks) as possible. Let... To perform secondary monitoring tasks The additional clinical value weights resulting from this are used to construct the following unified overall objective function:

[0080] ;

[0081] in, , It is an adjustable weighting parameter that quantifies the conversion relationship between clinical utility and energy consumption cost, and is used to reflect the preference between the two sub-goals of "energy saving" and "efficiency improvement". The value can be set according to the specific application scenario: for example, during the routine stable monitoring phase, a smaller value can be set. Prioritizing long battery life; however, when potential health risks are detected and more comprehensive diagnostic information is needed, the system or medical personnel can dynamically adjust the power level. This value incentivizes the system to expend more energy to activate all valuable monitoring tasks. Specifically, The value can be calibrated based on Pareto optimal frontier analysis or dynamically adjusted according to the remaining energy.

[0082] The method provided in this embodiment has at least the following beneficial effects:

[0083] This embodiment draws inspiration from the multi-commodity network flow problem. To address the optimal balance between the diverse needs of upper-level clinical monitoring tasks and the resource constraints of lower-level sensors, the convergence process of different types of physiological data from sensors to terminals is abstracted as a multi-commodity network flow problem. In scenarios involving multiple monitoring tasks, an objective function is established based on minimizing the difference between sensor energy consumption and the additional clinical value gains of secondary monitoring tasks. Corresponding overall constraints and decision variables are set, and a joint optimization problem is established by combining monitoring task decision-making with sensor resource allocation. By solving the joint optimization problem, joint optimization of monitoring task decision-making and sensor data resource allocation can be achieved, ensuring the effective execution of core monitoring tasks while maximizing operational efficiency and lifecycle.

[0084] In some embodiments of this application, step S300, which involves solving the joint optimization problem based on the overall constraints and the overall objective function to obtain the optimal solution for the decision variable terms, includes:

[0085] Step S310: Based on the Benders decomposition method, the joint optimization problem is divided into the main problem and sub-problems;

[0086] Step S320: Solve the main problem to estimate the binary decision variables corresponding to each of the multiple monitoring tasks; based on the estimated binary decision variables corresponding to each monitoring task, solve the sub-problems to evaluate the corresponding feasibility, and determine the cutting plane based on the feasibility, so as to optimize the estimated binary decision variables corresponding to each monitoring task according to the cutting plane, until the optimal solution of the decision variable terms is obtained.

[0087] In this embodiment, the joint optimization problem is a mixed-integer programming model that couples discrete task decision variables. and continuous network flow variables .

[0088] To achieve efficient solution, this embodiment uses the Benders decomposition method to decompose the joint optimization problem into a task selection master problem (TSMP) and a data aggregation subproblem (DASP).

[0089] TSMP is an integer programming problem used to determine which monitoring tasks to perform (i.e., solving for binary decision variables). This involves identifying a set of candidate task combinations. The detailed cost of data aggregation (sensor energy consumption) can be abstracted into a variable to be estimated, and this estimate can be continuously refined through the "cutting plane" constraints obtained by DASP iteration. The cutting plane refers to the constraint condition used to limit the feasible region of the main problem. Specifically, it can be implemented based on the dual variables of the subproblems or extreme ray generation. By continuously adding cutting planes, infeasible or suboptimal candidate task combinations can be gradually eliminated.

[0090] DASP receives and evaluates a combination of candidate tasks passed from TSMP. This combination is a predetermined decision vector, denoted as . ,in, Indicates monitoring task To be executed, and This indicates that the operation will not be performed. For a given vector... DASP is constructed as a linear programming (LP) problem to evaluate its feasibility and calculate its minimum total sensor energy consumption based on the objective function. Decision variables in the TSMP problem. In DASP, it is considered a known constant parameter. The results of DASP will provide crucial energy consumption feedback for monitoring task decisions.

[0091] Compared to existing technologies, traditional methods typically handle task scheduling and resource allocation separately, leading to suboptimal optimization results or high computational complexity. Existing technologies struggle to simultaneously handle discrete task decision variables and continuous resource allocation variables, easily getting trapped in local optima or failing to meet real-time requirements. This method decomposes the mixed-integer programming problem into independent problems using a decomposition optimization framework. Through this framework and a cutting plane iteration mechanism, it improves resource utilization efficiency in complex human monitoring scenarios. This method is particularly suitable for wearable device scenarios with limited medical resources, maximizing the clinical benefits of monitoring tasks within a limited energy budget.

[0092] In some embodiments of this application, step S320 involves solving the main problem to estimate the binary decision variables corresponding to each of the multiple monitoring tasks; based on the estimated binary decision variables corresponding to each monitoring task, solving sub-problems to assess the corresponding feasibility; and determining a cutting plane based on the feasibility to optimize the estimated binary decision variables corresponding to each monitoring task according to the cutting plane until the optimal solution for the decision variable terms is obtained, including:

[0093] Step S3210: Preset the optimality tolerance, the Benders cut plane set, the number of iterations, the upper bound, and the lower bound; wherein, the initial number of iterations is 0, the initial upper bound is positive infinity, the initial lower bound is negative infinity, and the initial Benders cut plane set is an empty set; the optimality tolerance refers to the allowed difference threshold between the upper and lower bounds, which serves as a termination condition for the algorithm. The Benders cut plane set refers to the accumulated set of constraints used to limit the feasible region of the main problem.

[0094] Step S3220: Perform iterative solution; any iteration process includes:

[0095] Step S3221: Based on the current set of Benders cutting planes and the objective function of the main problem, the branch and bound method is used to solve the main problem, estimating the binary decision variables corresponding to each of the multiple monitoring tasks and the current lower bound; based on the estimated binary decision variables corresponding to each of the multiple monitoring tasks, the current candidate task combination is obtained; the branch and bound method is an algorithm for solving integer programming problems, which can be implemented by constructing a search tree and pruning invalid branches, and is used to handle the binary decision variables in the main problem.

[0096] Step S3222: Based on the current candidate task combination and the objective function of the sub-problem, the augmented path method is used to solve the sub-problem, and the solution result of the sub-problem is obtained. The solution result of the sub-problem includes at least the feasibility of the current candidate task combination, the data transmission rate allocated by the sensor in the corresponding link for transmitting the corresponding first physiological data, the amount of data of the corresponding initial physiological data that the sensor locally simplifies, and the current total energy consumption of the sensor. The augmented path method is a solution algorithm for network flow problems. Specifically, it can be implemented by constructing a residual network and finding flow augmentation paths, and is used to handle continuous variable optimization in the sub-problem.

[0097] Step S3223: Based on the solution results of the subproblems, calculate the new Benders cut plane and the current upper bound, and update the Benders cut plane set according to the new Benders cut plane;

[0098] Step S3224: Based on the current upper bound, the current lower bound, and the optimality tolerance, determine whether convergence has occurred. If convergence has occurred, use the data transmission rate allocated by the sensor in the corresponding link for transmitting the corresponding first physiological data in the current candidate task combination and the solution results of the subproblems, as well as the amount of data of the corresponding initial physiological data that has been simplified locally by the sensor, as the optimal solution of the decision variable. If convergence has not occurred, continue to the next iteration.

[0099] Specifically, the main problem determines the binary decision variables for candidate task combinations, while sub-problems verify the feasibility of resource allocation for these combinations. In each iteration, after the main problem generates candidate task combinations using a branch-and-bound method, the sub-problems construct an augmented network graph containing virtual nodes to calculate link data transmission rates and reduce data volume. When a sub-problem is feasible, an optimality cut plane is generated; if infeasible, a feasibility cut plane is generated. These cut planes are fed back into the main problem to narrow the solution space. Through gradual approximation using upper and lower bounds, the optimal task combination and resource allocation scheme that satisfies resource constraints and optimizes energy consumption and revenue is ultimately obtained.

[0100] In some embodiments of this application, the objective function of the subproblem is a first function;

[0101] Based on the above embodiments: the goal of DASP is to combine candidate tasks. Find the lowest-cost data aggregation solution, i.e., minimize total energy consumption.

[0102] ;

[0103] The constraints of the sub-problem include: conditions four to seven, and condition eight, which is transformed from condition one; condition eight, the basis for the terminal to execute the current candidate task combination is: the data flow of each type of first physiological data required for the terminal to receive the current candidate task combination is not less than the preset minimum requirement;

[0104] The constraints of DASP originate from the original joint optimization problem model and are based on the combination of input candidate tasks. Perform parameterization:

[0105] Condition 8 (C8): This constraint is crucial in connecting the main problem decision with the subproblems. For any first physiological data... Its terminal of All must be satisfied in The demand for activated monitoring tasks. Therefore, the total demand is... .

[0106] ;

[0107] in, .

[0108] The decision variables for the sub-problem include: the data transmission rate allocated by the sensor in the corresponding link for transmitting the corresponding first physiological data. And the amount of data corresponding to the initial physiological data that was simplified locally by the sensor. ;

[0109] Step S3222 uses the augmented path method to solve the subproblem and obtains the solution results for the subproblem, including:

[0110] Step S410: Construct an original network topology map based on multiple sensors and terminals; the nodes in the original network topology map are multiple sensors and terminals, and the edges are the link relationships between multiple sensors and terminals.

[0111] Step S420: Based on the original network topology and the current candidate task combination, determine the augmented network graph and the residual network graph. The augmented network graph is a network topology formed by adding virtual source nodes and virtual sink nodes and their corresponding connections to the original network topology. The residual network graph has the same nodes as the augmented network graph, and the edge set and edge attributes of the residual network graph are related to the data traffic status in all links.

[0112] To efficiently solve the above DASP problem, this method transforms the original problem into a standard single-source, single-sink minimum cost flow problem, specifically including:

[0113] Building augmenting networks Based on the constraints of DASP, the original multi-source (sensor) ), Single Exchange (Terminal) ) network Transformed into a single-source, single-sink augmented network diagram Specifically:

[0114] Step S4211, introduce the virtual total source point and virtual hub .

[0115] Step S4212, for each that can generate the first physiological data Sensors Add a line from arrive The virtual source edge, whose capacity represents the sensor The maximum supply, i.e., the amount of raw physiological data. .

[0116] Step S4213, for the terminal h Each type of primary physiological data that needs to be collected Add a line from h arrive The virtual sink edge has a capacity equal to the total demand for the first physiological data: .

[0117] Step S4214, Physical Link In augmented network diagram The total capacity is reserved in the middle. The cost of each edge in the augmented network graph is determined by the corresponding energy consumption coefficient (e.g., ...). Configure it.

[0118] Constructing the residual network graph Residual network diagram With augmented network diagram They share the same set of nodes, but their edge sets are different. The properties of the edges (residual capacity, cost) depend on the flow state. For augmented network graphs any edge in Its capacity is The cost is Current traffic is .

[0119] Step S4221, if Then in the residual network graph There exists a forward edge in the middle. Its residual capacity The cost is This means that the flow rate can be increased in the original direction.

[0120] Step S4222, if Then in the residual network graph There exists a backward edge Its residual capacity The cost is This means that traffic can be reversed or pushed back to find a better path.

[0121] Step S4223: At the start of the iteration, all flows are initialized to 0 (i.e., ... ), initial residual network graph Structure and augmented network diagram Similarly, all edges are oriented in the same direction, the residual capacity equals its original capacity, and the cost equals the original cost. Subsequent augmenting path searches will be performed on this initial residual network graph.

[0122] Step S430: Based on the residual network graph, the virtual source point, the virtual sink point, and the amount of data required to be transferred from the virtual source point to the virtual sink point based on the current candidate task combination, calculate the solution result of the subproblem.

[0123] To efficiently solve the minimum-cost network flow problem described by DASP, this embodiment gradually allocates traffic by repeatedly searching for and utilizing the lowest-cost path in the network until all task requirements are met or the problem is deemed infeasible. This method avoids the high computational overhead of general linear programming solvers.

[0124] Step S4310, set the total flow rate Residual network diagram Perform initialization.

[0125] Step S4320, when the total flow Repeat the following steps:

[0126] Step S4321, Search for the minimum cost augmenting path: in the current residual network graph In the process, using the energy cost per unit of traffic as the edge weight, a path search is performed to find the path from the virtual origin. To the virtual hub The lowest cost path Shortest path algorithms that can handle negative weights can be selected, such as the Bellman-Ford algorithm. Bellman-Ford can guarantee that it can correctly find the minimum cost path when there are negative weights but no negative weight loops.

[0127] Step S4322: Determine the validity of the path;

[0128] If in the residual network graph There is no such thing as from arrive This path means that the currently allocated traffic cannot be increased, but the total demand... If the condition is not yet met, the algorithm terminates, determining that the DASP problem is valid for the given condition. It's not feasible. If a path could be found... Then proceed to the next step.

[0129] Step S4323: Calculate the path bottleneck capacity and amplify the flow;

[0130] First, calculate the path. bottleneck capacity ,in The edges on the path are in the residual network graph The remaining capacity in. Then, along that path. Traffic augmentation refers to the process of increasing traffic on the network. Update: For the path Each forward edge on Update its traffic to As for the path Each backward edge on Then update the flow of its corresponding primary edge to Finally, update the total flow that has been satisfied to... .

[0131] Step S4324, update the residual network;

[0132] Based on traffic Changes, recalculate and update the residual network graph The residual capacity of all affected edges and the possible new backward edges.

[0133] Step S4330, the iteration loop ends (i.e.) At this point, the data requirements for all tasks have been met. The algorithm terminates, determining that the DASP problem is feasible. Current... This is the optimal solution. Then, the total energy cost is calculated using the objective function. .

[0134] Compared to existing technologies, which typically employ static resource allocation methods and cannot dynamically adapt to changes in data transmission requirements arising from different combinations of candidate tasks, this method achieves cross-link global resource coordination and allocation by constructing an augmented network graph and a dynamically updated residual network graph. Furthermore, it introduces virtual source and sink points, simplifying the modeling complexity of multi-link collaborative transmission and enabling the algorithm to efficiently handle resource allocation problems in large-scale sensor networks.

[0135] In some embodiments of this application, step S3223, calculating the new Benders cut plane based on the solution results of the sub-problem, includes:

[0136] Step S410: If the subproblem has a feasible solution, calculate the optimal cutting plane based on a set of optimal dual variables determined by the constraints of the subproblem, the current combination of candidate tasks, and the current total energy consumption of the sensor.

[0137] When DASP is feasible, it is necessary to And the decisions that lead to the monitoring task Establish mathematical relationships to provide cost gradient information for the main problem.

[0138] A feasible solution to DASP corresponds to a set of optimal dual variables. These correspond to the five constraints of the aforementioned subproblems. According to the strong duality theorem, if a linear programming problem (the primal problem) has an optimal solution, then its dual problem must also have an optimal solution, and the objective function values ​​corresponding to these two optimal solutions are equal. Using this property, an optimality cutting plane is constructed, expressed by the following formula:

[0139] ;

[0140] in, This is an auxiliary decision variable (i.e., a preset minimum total energy consumption) introduced in the TSMP problem, used to represent the estimate of the minimum total power consumption for the DASP problem. It is obtained through the Benders optimality cutting plane. The value is constantly constrained, thus approximating the true minimum energy cost. It is the optimal dual variable corresponding to constraint C1.

[0141] This inequality can provide a lower bound for linearity.

[0142] Step S420: When there is no feasible solution to the subproblem, calculate the feasible cutting plane based on the extreme rays corresponding to a set of optimal dual variables.

[0143] The infeasibility of the DASP problem corresponds to the unboundedness of its dual problem. When the total requirement is not satisfied, a set of dual extreme rays can be found. Using this ray, a feasible cutting plane can be constructed, and its formula is expressed as follows:

[0144] ;

[0145] in, It is a constant calculated from the extreme rays and the constraint right-hand side. This inequality can effectively eliminate a series of infeasible task combinations, including the currently infeasible solution.

[0146] Compared to existing technologies, traditional Benders decomposition methods typically generate only a single type of cutting plane when dealing with mixed-integer programming problems, failing to distinguish between the quality differences of feasible solutions and the root causes of infeasible solutions. This method dynamically determines the feasibility status of subproblems and generates two types of cutting planes accordingly. These two types work synergistically, simultaneously achieving effective pruning of the solution space and gradual improvement of solution quality during the optimization process. This ensures balanced energy consumption of sensors and reasonable allocation of terminal computing resources.

[0147] In some embodiments of this application, the objective function of the main problem is to minimize the difference between the preset minimum total energy consumption and the second function;

[0148] Based on the above embodiments, the main problem is an integer linear programming problem, the goal of which is to make the optimal choice of monitoring tasks while satisfying all known constraints (including the feedback cut plane from the first part).

[0149] The objective function of the main problem:

[0150] ;

[0151] The constraints of the main problem include:

[0152] C2: ;

[0153] C3: ;

[0154] Condition 9 (C9) Benders Feedback Constraint: During the iteration of step S3220, the main problem must satisfy all cumulative cutting plane constraints fed back by subproblems. These constraints are divided into two cases based on the solution state of the subproblems:

[0155] When the subproblem has a feasible solution, for each iteration (round index is...) The following optimality cutting plane constraints must be satisfied:

[0156] ;

[0157] in, It is the set of iterations in which all subproblems have feasible solutions. It is the first The main problem of the round provides a combination of candidate tasks. It is the optimal dual variable obtained by solving the wheel problem.

[0158] This constraint is the minimum energy cost. It provides a dynamically updated lower bound.

[0159] When the subproblem has no feasible solution, for each iteration (round index is...) The following feasibility cutting plane constraints must be met:

[0160] ;

[0161] in, It is the set of iterations for all subproblems that have no feasible solutions. and The parameters are calculated based on the dual extreme rays discovered in this round. This constraint is used to remove invalid combinations of tasks from the feasible region of the main problem that have been shown to lead to unsatisfactory resources.

[0162] Step S3221, which estimates the binary decision variables for each of the multiple monitoring tasks, includes:

[0163] Step S510: Construct a branch and bound tree; the nodes of the branch and bound tree represent subproblems.

[0164] The core of branch and bound is to obtain the bound of the objective value by solving a relaxed version of the problem, so as to guide the search direction and pruning.

[0165] Step S520: For each node in the branch and bound tree, the subproblem is transformed into an LP relaxation problem.

[0166] Step S530: Solve the LP relaxation problem based on the current Benders cut plane set to obtain the solution to the LP relaxation problem and the current lower bound obtained based on the solution to the LP relaxation problem.

[0167] In each node of the branch and bound tree, the integer constraints of the TSMP problem are applied. relaxation This yields an LP relaxation problem. Due to the existence of the Benders cut plane, the solution to this relaxation problem is... It can provide a tight lower bound for the current branch.

[0168] Step S540: Based on the solution to the LP relaxation problem and the clinical value weights corresponding to each monitoring task, estimate the binary decision variables corresponding to each monitoring task.

[0169] Check LP relaxation .like If the integer is true, then a feasible solution is obtained. If a component exists... If the value is a score, then that variable is selected for branching. This embodiment employs a utility-first strategy: prioritizing the weight with the highest clinical value. fractional variables Create a branch, add two child nodes, and add constraints to each. and .

[0170] Specifically, the solution to the main problem unfolds by constructing a branch-and-bound tree, where each tree node corresponds to an LP relaxation version of a subproblem. During node processing, the subproblem is first transformed into an LP relaxation problem, eliminating the integer constraints on the binary decision variables and making them continuous variables. Then, combining the Benders cut plane constraints accumulated in the current iteration, the linear programming solver is invoked to calculate the solution to the relaxation problem, and the current lower bound is updated based on the objective function value of the solution. Based on the continuous values ​​of the binary decision variables in the relaxation solution and their corresponding clinical value weights, it is estimated whether each monitoring task should be selected for execution.

[0171] This method can dynamically balance sensor energy consumption and the clinical benefits of secondary tasks while ensuring the execution of core monitoring tasks, thereby optimizing the overall system efficiency.

[0172] In some embodiments, pruning can be performed in a timely manner to avoid exploring hopeless search spaces. For example, when expanding a new node, a pruning decision is made: if the node's LP relaxation problem has no solution, then the branch is pruned; if the node's lower bound is not met, then the branch is pruned. If the target value is greater than or equal to the currently found optimal integer solution, then prune the branch; if the node produces an integer feasible solution, then update the global optimal solution and prune the branch.

[0173] For ease of understanding, this application provides an embodiment of a joint optimization method for a vital signs monitoring task, which includes two parts:

[0174] The first part constructs a joint optimization problem model, formalizing the home monitoring scenario into a unified mathematical optimization model:

[0175] (1) Scene by A set of sensors A gateway (i.e., the terminal) and a remote medical cloud platform (i.e., server) configuration. Sensors generate different types of physiological data. At the gateway The data is collected and used to perform composite monitoring tasks.

[0176] (2) Transfer data from the sensor to the gateway The convergence process can be abstracted into a multi-product network flow problem.

[0177] The model includes flow conservation constraints (constraint C6), link capacity constraints (constraint C7), and sensor local data processing constraints (constraint C5). Decision variables include: and .

[0178] (3) Introduce binary decision variables Indicates monitoring task Whether to execute. The core constraints include: the data traffic required for task execution must meet the minimum requirements (i.e., constraint C1), the core monitoring task must be executed (i.e., constraint C2), and the total computational overhead of all executed tasks cannot exceed the gateway's processing capacity limit (i.e., constraint C3).

[0179] (4) Quantify the energy consumption of each sensor and stipulate that its total energy consumption shall not exceed its respective energy budget limit (i.e., constraint C4).

[0180] (5) The final problem is formalized as finding an optimal combination of decision variables. To minimize an objective function that combines total energy consumption with weighted clinical utility:

[0181] ;

[0182] The following describes the process of solving the problem model:

[0183] Step S910: Set the number of iterations Optimal tolerance Initialize the Benders cut plane set to an empty set.

[0184] Output: Number of iterations , Benders cutting plane set Upper Realm The lower realm .

[0185] Step S920: Solve the main problem to obtain a candidate task combination. .

[0186] Input: The current TSMP (containing the cut plane set) );

[0187] Output: Candidate task combinations The current lower bound ;

[0188] Step S930: with Given the input, solve the data aggregator problem.

[0189] Input: Candidate task combinations ;

[0190] Output: Solution results for the DASP problem (feasibility, minimum energy cost) (dual variables, etc.)

[0191] Step S940: Based on the solution results of DASP, generate a new optimality cut plane or feasibility cut plane and add it to the Benders cut plane set.

[0192] Input: The solution to the DASP problem;

[0193] Output: New Benders cut planes, updated set of cut planes The updated upper realm ;

[0194] Step S950 (Convergence Judgment): Calculate the difference between the current upper bound and the current lower bound, i.e. If the difference is less than the preset optimal tolerance... If the algorithm converges, it will output the current optimal task combination. Otherwise, let Return to step S920 and begin a new round of iterations.

[0195] Input: upper bound The lower realm Optimal tolerance ;

[0196] Output: Optimal task combination Or a signal to continue iterating.

[0197] This method has at least the following beneficial effects:

[0198] This method draws inspiration from the multi-commodity network flow problem. To address the optimal balance between the diverse needs of upper-level clinical monitoring tasks and the resource constraints of lower-level sensors, it abstracts the convergence process of different types of physiological data from sensors to terminals as a multi-commodity network flow problem. In scenarios involving multiple monitoring tasks, an objective function is established based on minimizing the difference between sensor energy consumption and the additional clinical value gains of secondary monitoring tasks. Corresponding overall constraints and decision variables are set, and a joint optimization problem is established by combining monitoring task decision-making with sensor resource allocation. By solving the joint optimization problem, the joint optimization of monitoring task decision-making and sensor data resource allocation can be achieved, ensuring the effective execution of core monitoring tasks while maximizing operational utility and lifecycle.

[0199] This method divides the original joint optimization problem into two core problems: resource cost assessment and task selection optimization. Through iterative interaction between the two problems, the solution to the problem defined by the original problem model is completed collaboratively, which can efficiently solve the optimal task combination under all physical and clinical constraints.

[0200] like Figure 2 One embodiment of this application provides a joint optimization system for vital sign monitoring tasks, the system comprising:

[0201] Human body monitoring platform 1100; Human body monitoring platform 1100 includes multiple sensors 1110 and terminals 1120;

[0202] Server 1200, the server includes:

[0203] Problem building module 1210 is used to classify multiple monitoring tasks into core monitoring tasks and secondary monitoring tasks;

[0204] The objective function determination module 1220 is used to determine the overall constraints and overall objective function of the joint optimization problem. The overall objective function is to minimize the difference between the first function and the second function. The first function includes the idle operating energy consumption of the sensor, the computational energy consumption of the sensor performing local simplification, and the transmission energy consumption of the sensor transmitting the corresponding first physiological data to the terminal. The transmission energy consumption is related to the data transmission rate allocated by the sensor on the corresponding link for transmitting the corresponding first physiological data. The second function includes the additional clinical value benefit of performing secondary monitoring tasks. The overall constraints include:

[0205] Condition 1: The basis for the terminal to execute monitoring tasks is that the data flow of each type of first physiological data required for the monitoring task received by the terminal is not less than the preset minimum requirement; Condition 2: All core monitoring tasks must be executed; Condition 3: The total computing resources of the terminal to execute at least one monitoring task do not exceed the terminal's maximum computing resources; Condition 4: The total energy consumption of each sensor does not exceed the sensor's energy consumption limit; where the total energy consumption of the sensor is equal to the sum of the corresponding idle operation energy consumption, computing energy consumption, and transmission energy consumption; Condition 5: The first physiological data after local simplification by the sensor does not exceed the corresponding initial physiological data; Condition 6: The data flow transmitted by the sensor on the corresponding link and the data flow received by the terminal on the corresponding link are conserved; Condition 7: The data flow transmitted in any link does not exceed the maximum capacity of the corresponding link;

[0206] The problem-solving module 1230 is used to determine the decision variables of the joint optimization problem and solve the joint optimization problem based on the overall constraints and the overall objective function to obtain the optimal solution of the decision variables. The decision variables of the joint optimization problem include: binary decision variables, the data transmission rate allocated by the sensor in the corresponding link for transmitting the corresponding first physiological data, and the amount of data of the corresponding initial physiological data that is locally simplified by the sensor. The binary decision variables use 0 or 1 to represent whether each monitoring task in the multiple monitoring tasks is executed.

[0207] It should be noted that the joint optimization system for vital sign monitoring tasks provided in this embodiment and the joint optimization method for vital sign monitoring tasks described above are based on the same inventive concept. Therefore, the content of the joint optimization system for vital sign monitoring tasks described in this embodiment is also applicable to the content of the joint optimization method for vital sign monitoring tasks described above, and will not be repeated here.

[0208] like Figure 3 An embodiment of this application also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the aforementioned joint optimization method for a vital sign monitoring task. The electronic device includes:

[0209] At least one battery;

[0210] At least one memory;

[0211] At least one processor;

[0212] At least one program;

[0213] The program is stored in memory, and the processor executes at least one program to implement the joint optimization method for a vital signs monitoring task described above in this disclosure.

[0214] This electronic device can be any smart terminal, including mobile phones, tablets, personal digital assistants (PDAs), and in-vehicle computers.

[0215] The electronic devices according to embodiments of this application will now be described in detail.

[0216] The processor 1600 can be implemented using a general-purpose central processing unit (CPU), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this disclosure.

[0217] The memory 1700 can be implemented as a read-only memory (ROM), static storage device, dynamic storage device, or random access memory (RAM). The memory 1700 can store the operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 1700 and is called by the processor 1600 to execute a joint optimization method for a vital sign monitoring task according to an embodiment of this disclosure.

[0218] The input / output interface 1800 is used to implement information input and output.

[0219] The communication interface 1900 is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.).

[0220] Bus 2000 transmits information between various components of the device (e.g., processor 1600, memory 1700, input / output interface 1800, and communication interface 1900);

[0221] The processor 1600, memory 1700, input / output interface 1800 and communication interface 1900 are connected to each other within the device via bus 2000.

[0222] This disclosure also provides a storage medium, which is a computer-readable storage medium storing computer-executable instructions for causing a computer to perform the above-described actions.

[0223] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0224] The embodiments described in this disclosure are for the purpose of more clearly illustrating the technical solutions of this disclosure and do not constitute a limitation on the technical solutions provided by this disclosure. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by this disclosure are also applicable to similar technical problems.

[0225] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of this disclosure, and may include more or fewer steps than shown, or combine certain steps, or different steps.

[0226] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0227] Those skilled in the art will understand that all or some of the steps in the methods disclosed above, as well as the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, or suitable combinations thereof.

[0228] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0229] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0230] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0231] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0232] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0233] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes multiple instructions to cause an electronic device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing programs, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0234] The above is a detailed description of the preferred embodiments of this application. However, the embodiments of this application are not limited to the above-described implementation methods. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the embodiments of this application. All such equivalent modifications or substitutions are included within the scope defined by the claims of the embodiments of this application.

Claims

1. A method for joint optimization of a vital sign monitoring task, characterized in that, The application is applied to a human body monitoring platform with multiple monitoring tasks, and the human body monitoring platform comprises: a plurality of sensors, which can collect corresponding raw physiological data from a human body, perform local simplification on the raw physiological data to obtain first physiological data, and transmit the first physiological data to a terminal; a terminal connected with the plurality of sensors through corresponding links, which can execute at least one monitoring task in the multiple monitoring tasks, and one monitoring task needs to be cooperatively executed through multiple first physiological data; a joint optimization method of the vital sign monitoring task comprises: dividing the multiple monitoring tasks into core monitoring tasks and secondary monitoring tasks; determining total constraint conditions and a total objective function of a joint optimization problem; the total objective function is to minimize the difference between a first function and a second function, the first function includes idle running energy consumption of a sensor, calculation energy consumption of the sensor for local simplification, and transmission energy consumption of the sensor for transmitting corresponding first physiological data to a terminal; the transmission energy consumption is associated with a data transmission rate allocated by the sensor on a corresponding link for transmitting corresponding first physiological data; the second function includes additional clinical value benefits of executing secondary monitoring tasks; the total constraint conditions include: condition one, a basis for the terminal to execute a monitoring task is that data traffic of each kind of first physiological data required by the monitoring task received by the terminal is not lower than a preset minimum requirement; condition two, all core monitoring tasks need to be executed; condition three, total calculation resources of the terminal for executing at least one monitoring task do not exceed maximum calculation resources of the terminal; condition four, total energy consumption of each sensor does not exceed an upper limit of energy consumption of the sensor; wherein the total energy consumption of the sensor is equal to the sum of corresponding idle running energy consumption, calculation energy consumption and transmission energy consumption; condition five, first physiological data simplified locally by the sensor does not exceed corresponding initial physiological data; condition six, data traffic transmitted on a corresponding link by the sensor and data traffic received on the corresponding link by the terminal is conserved; condition seven, data traffic transmitted in any link does not exceed maximum capacity of the corresponding link; determining decision variable items of the joint optimization problem, and solving the joint optimization problem based on the total constraint conditions and the total objective function to obtain an optimal solution of the decision variable items; the decision variable items of the joint optimization problem include binary decision variables, data transmission rates allocated by the sensor on corresponding links for transmitting corresponding first physiological data, and data amounts of corresponding initial physiological data discarded by the sensor for local simplification; the binary decision variables represent whether each monitoring task in the multiple monitoring tasks is executed by using 0 or 1.

2. The method of joint optimization of a sign monitoring task according to claim 1, wherein, The second function is a sum of products of binary decision variables corresponding to secondary monitoring tasks and additional clinical value weights.

3. The method of joint optimization of a sign monitoring task according to claim 2, characterized in that, The solving of the joint optimization problem based on the total constraint conditions and the total objective function to obtain the optimal solution of the decision variable items comprises: partitioning the joint optimization problem into a master problem and a sub-problem based on the Benders decomposition method, wherein the master problem is an integer programming problem and the sub-problem is a linear programming problem, the objective function of the master problem is associated with the second function, and the objective function of the sub-problem is associated with the first function; solving the master problem to estimate the binary decision variable corresponding to each monitoring task in the plurality of monitoring tasks, and solving the sub-problem based on the estimated binary decision variable corresponding to each monitoring task to evaluate the corresponding feasibility, and determining a cut plane based on the feasibility to optimize the estimated binary decision variable corresponding to each monitoring task according to the cut plane until an optimal solution of the decision variable term is obtained.

4. The method of joint optimization of a vital sign monitoring task according to claim 3, characterized in that, The solving the master problem to estimate the binary decision variable corresponding to each monitoring task in the plurality of monitoring tasks; solving the sub-problem based on the estimated binary decision variable corresponding to each monitoring task to evaluate the corresponding feasibility; determining a cut plane based on the feasibility to optimize the estimated binary decision variable corresponding to each monitoring task according to the cut plane until an optimal solution of the decision variable term is obtained, comprising: presetting an optimality tolerance, a Benders cut plane set, an iteration number, an upper bound and a lower bound, wherein the initial iteration number is 0, the initial upper bound is positive infinity, the initial lower bound is negative infinity, and the initial Benders cut plane set is an empty set; the optimality tolerance serves as a threshold; performing iterative solving, wherein any iteration process comprises: solving the master problem by using a branch and bound method according to the current Benders cut plane set and the objective function of the master problem to estimate the binary decision variable corresponding to each monitoring task in the plurality of monitoring tasks and the current lower bound, and obtaining a current candidate task combination according to the estimated binary decision variable corresponding to each monitoring task in the plurality of monitoring tasks; solving the sub-problem by using an augmented path method according to the current candidate task combination and the objective function of the sub-problem to obtain a solving result of the sub-problem; the solving result of the sub-problem at least comprises the feasibility of the current candidate task combination, the data transmission rate allocated by the sensor to the corresponding first physiological data in the corresponding link for transmission, the data amount of the corresponding initial physiological data that is locally simplified by the sensor, and the current total energy consumption of the sensor; calculating a new Benders cut plane and the current upper bound according to the solving result of the sub-problem, and updating the Benders cut plane set according to the new Benders cut plane; judging whether to converge based on the current upper bound, the current lower bound and the optimality tolerance; in the case of convergence, taking the current candidate task combination, the data transmission rate allocated by the sensor to the corresponding first physiological data in the corresponding link for transmission, and the data amount of the corresponding initial physiological data that is locally simplified by the sensor in the solving result of the sub-problem as the optimal solution of the decision variable term; in the case of non-convergence, continuing the next round of iteration.

5. The method of joint optimization of a vital sign monitoring mission according to claim 4, characterized in that, The objective function of the sub-problem is the first function. The constraint conditions of the sub-problem include: conditions four to seven, and condition eight transformed based on condition one; condition eight includes that the terminal performs the current candidate task combination based on that the data flow of each first physiological data required by the terminal receiving the current candidate task combination is not lower than a preset minimum requirement; The decision variables of the sub-problem include: the data transmission rate of the sensor in the corresponding link for transmitting the corresponding first physiological data, and the data amount of the corresponding initial physiological data that is locally pruned by the sensor; The sub-problem is solved by using the augmented path method to obtain a solution of the sub-problem, which includes: An original network topology graph is constructed based on the plurality of sensors and the terminal; nodes in the original network topology graph are the plurality of sensors and the terminal, and edges are link relationships between the plurality of sensors and the terminal; An augmented network graph and a residual network graph are determined according to the original network topology graph and the current candidate task combination; the augmented network graph is formed by adding a virtual total source point, a virtual total sink point and corresponding connections to the original network topology graph; the residual network graph has the same nodes as the augmented network graph, and the edge set and the attribute of the edge of the residual network graph are related to the data flow state in all links; The solution of the sub-problem is calculated according to the residual network graph, the virtual total source point, the virtual total sink point, and the data demand amount required to be transmitted from the virtual total source point to the virtual total sink point based on the current candidate task combination.

6. The method of joint optimization of a vital sign monitoring task according to claim 5, wherein, The new Benders cut plane is calculated according to the solution of the sub-problem, which includes: In the case that the sub-problem has a feasible solution, an optimality cut plane is calculated according to a set of optimal dual variables determined based on the constraint conditions of the sub-problem, the current candidate task combination, and the current total energy consumption of the sensor; In the case that the sub-problem has no feasible solution, a feasibility cut plane is calculated according to an extreme ray corresponding to the set of optimal dual variables.

7. The method of joint optimization of a vital sign monitoring task according to claim 6, characterized in that, The objective function of the main problem is to minimize the difference between the preset minimum total energy consumption and the second function; The constraint conditions of the main problem include: condition two, condition three, and condition nine; condition nine includes that, in the case that the sub-problem has a feasible solution, each iteration satisfies the optimality cut plane constraint, and in the case that the sub-problem has no feasible solution, each iteration satisfies the feasibility cut plane constraint; The binary decision variable corresponding to each monitoring task in the plurality of monitoring tasks is estimated, which includes: A branch and bound tree is constructed; nodes of the branch and bound tree represent the sub-problem; The sub-problem is converted into an LP relaxation problem in the nodes of the branch and bound tree; The LP relaxation problem is solved according to the current set of Benders cut planes to obtain a solution of the LP relaxation problem and a current lower bound obtained based on the solution of the LP relaxation problem; The binary decision variable corresponding to each monitoring task is estimated according to the solution of the LP relaxation problem and the clinical value weight corresponding to each monitoring task.

8. A system for joint optimization of vital sign monitoring tasks, characterized in that, The system includes: The human body monitoring platform comprises a plurality of sensors and a terminal. The sensors can collect corresponding raw physiological data from the human body, perform local simplification on the raw physiological data to obtain first physiological data, and transmit the first physiological data to the terminal. The terminal is connected to the plurality of sensors through corresponding links. The terminal can execute at least one monitoring task in a plurality of monitoring tasks, and one monitoring task needs to be cooperatively executed through a plurality of first physiological data. A server, comprising: a problem building module configured to divide the plurality of monitoring tasks into core monitoring tasks and secondary monitoring tasks; a target function determining module configured to determine total constraint conditions and a total target function of a joint optimization problem; the total target function is to minimize the difference between a first function and a second function; the first function comprises idle running energy consumption of the sensors, calculation energy consumption of the sensors for local simplification, and transmission energy consumption of the sensors for transmitting corresponding first physiological data to the terminal; the transmission energy consumption is associated with a data transmission rate allocated by the sensors on the corresponding link for transmitting the corresponding first physiological data; the second function comprises additional clinical value benefits of executing the secondary monitoring tasks; the total constraint conditions comprise: condition one, a basis for the terminal to execute the monitoring tasks is that data traffic of each kind of first physiological data required by the monitoring tasks received by the terminal is not lower than a preset minimum requirement; condition two, all core monitoring tasks need to be executed; condition three, total calculation resources of the terminal for executing at least one monitoring task are not more than maximum calculation resources of the terminal; condition four, total energy consumption of each sensor is not more than an upper limit of energy consumption of the sensor; wherein the total energy consumption of the sensor is equal to the sum of the corresponding idle running energy consumption, calculation energy consumption and transmission energy consumption; condition five, the first physiological data after local simplification of the sensor is not more than corresponding initial physiological data; condition six, data traffic transmitted by the sensor on the corresponding link and data traffic received by the terminal on the corresponding link is conserved; condition seven, data traffic transmitted in any link is not more than maximum capacity of the corresponding link; a problem solving module configured to determine decision variable items of the joint optimization problem, and solve the joint optimization problem based on the total constraint conditions and the total target function to obtain an optimal solution of the decision variable items; the decision variable items of the joint optimization problem comprise binary decision variables, data transmission rates allocated by the sensors on the corresponding links for transmitting the corresponding first physiological data, and data amounts of the corresponding initial physiological data discarded by the sensors for local simplification; the binary decision variables represent whether each monitoring task in the plurality of monitoring tasks is executed by using 0 or 1.

9. An electronic device, comprising: At least one controller and a memory connected in communication with the controller; the memory stores instructions executable by the at least one controller, and the instructions are executed by the at least one controller to enable the at least one controller to execute a joint optimization method of a sign monitoring task according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer readable storage medium stores computer executable instructions for causing a computer to perform the joint optimization method of a vital sign monitoring task according to any one of claims 1 to 7.

Images