A pre-hospital care helicopter optimal deployment method fusing weighted K-means and IABC

Abstract

The application discloses a pre-hospital care helicopter optimization deployment method fusing weighted K-means and IABC, first, patient data is acquired and the geographical position coordinates thereof are standardized pretreated, the patients are classified according to the injury degree and corresponding weight values are given, and a weighted data set is formed; subsequently, a weighted clustering square error and a silhouette coefficient are calculated, the elbow rule and the silhouette coefficient method are cooperatively decided, and the optimal helicopter deployment number k is determined; then, the weighted data set is taken as input, and the initial deployment position coordinates of the k helicopters are solved through a weighted K-means algorithm; then, a single-target fitness function integrated by three sub-functions including service coverage, rescue response time and economic cost is constructed, and a multi-target optimization problem is integrated into a single-target optimization problem; finally, an improved artificial bee colony (IABC) algorithm with a directional learning mechanism is adopted, the initial deployment position of the helicopter is taken as an initialization population for iterative updating, and the final pre-hospital care helicopter deployment coordinates are output.

Description

Technical Field

[0001] This invention relates to a method for optimizing the deployment of pre-hospital emergency helicopters by integrating weighted K-means and IABC. Background Technology

[0002] In modern public safety and emergency management systems, pre-hospital emergency care plays a crucial role in safeguarding the "golden hour" of life, and its response efficiency directly determines patients' cure and survival rates. With accelerated urbanization, increasingly complex transportation networks, and rising public demand for medical rescue services, the traditional ground ambulance system is increasingly facing timeliness bottlenecks when dealing with long-distance transport, traffic congestion, special terrain, and large-scale emergencies. Against this backdrop, air medical rescue, with its unique advantages of speed, flexibility, and less susceptibility to ground conditions, has become an indispensable component of modern pre-hospital emergency care systems.

[0003] However, building an efficient and reliable helicopter air ambulance network is not simply a matter of purchasing a fleet of advanced helicopters. Scientific planning and optimized deployment are the cornerstones determining the network's service effectiveness, response speed, and operational economics. An irrational layout will directly lead to numerous serious problems: creating helicopter service coverage blind spots, preventing some areas from receiving equitable emergency medical services; prolonging response times, causing patients to miss the optimal treatment window; and resulting in the idleness of limited helicopters and supporting resources, leading to low efficiency in public financial investment.

[0004] Therefore, how to scientifically and rationally deploy the number and location of pre-hospital emergency helicopters under limited budget constraints, in order to achieve the best balance between maximizing service coverage, minimizing rescue response time, and optimizing economic costs, has become a highly challenging practical problem. Summary of the Invention

[0005] This invention provides a pre-hospital emergency helicopter deployment optimization method that integrates weighted K-means and IABC to address the problems existing in the prior art.

[0006] The technical solutions adopted in this invention are as follows:

[0007] A method for optimizing the deployment of pre-hospital emergency helicopters by integrating weighted K-means and IABC includes the following steps:

[0008] S1: Acquire patient data and their geographic coordinates, and perform preprocessing;

[0009] S2: Based on the severity of the patient's injury, a weighted dataset is formed by classifying the injury and assigning corresponding weights.

[0010] S3: Calculate the weighted cluster squared error and the silhouette coefficient, and determine the optimal number of helicopters k based on the elbow rule and the silhouette coefficient method as a collaborative decision-making mechanism.

[0011] S4: Using a weighted dataset as input, solve for the initial deployment coordinates of k helicopters using a weighted K-means algorithm;

[0012] S5: Construct a single-objective fitness function that integrates three sub-functions: service coverage, rescue response time, and economic cost;

[0013] S6: An improved artificial bee colony algorithm with an introduced directional learning mechanism is adopted. The initial deployment position of the helicopter is used as the initial population for iterative updates, and the final deployment coordinates of the pre-hospital emergency helicopter are output.

[0014] Furthermore, in S1, the patient's geographic location coordinates are standardized and preprocessed using the following formula:

[0015] ,

[0016] ,

[0017] in, Indicates the first The geographical coordinates of each patient This represents the standardized geographic coordinates of the patient. and These represent the minimum and maximum values ​​of the geographic coordinates of all patients, respectively.

[0018] Furthermore, in S2, the weighting of the grading involves classifying patients into several grades based on the severity of their injury and assigning corresponding weight values ​​to different grades. ;

[0019] The obtained weight values ​​are used as the third column of data and combined with the standardized geographic coordinates to form a weighted dataset. , This indicates the number of patient data points.

[0020] Furthermore, in S3, the formula for calculating the weighted cluster squared error sum is:

[0021] ,

[0022] in, This indicates the number of pre-hospital emergency helicopters deployed. , Represents all cluster subsets, For the patient's geographical location, For the first Cluster centers, The patient To the cluster center The square of the distance;

[0023] The formula for calculating the profile coefficient is:

[0024] ,

[0025] ,

[0026] in, Indicates the first Profile coefficient of each patient It is the first The average dissimilarity between a patient and other patients within the same cluster subset It is the first Minimum average dissimilarity between a patient and other cluster subsets It is the average of the contour coefficients of all patients.

[0027] Furthermore, the optimal number of helicopters to deploy By searching The steepest point of descent and profile coefficient of the value The local maximum point is determined by collaborative decision-making.

[0028] Furthermore, in S4, the weighted K-means algorithm uses the weight values ​​in the weighted dataset. Adjust the contribution of each patient's data to the cluster centers, i.e., the weight value. The higher the value, the greater the impact of the corresponding patient data on the clustering process; The initial deployment coordinates of the helicopters are represented as follows: ,Right now The coordinates of the cluster centers.

[0029] Furthermore, in S5, the specific form of the integrated single-objective fitness function is:

[0030] ,

[0031] in, , and These represent the three sub-objective functions: service coverage, rescue response time, and economic cost, respectively. , and The weights represent the proportion of influence of the sub-objective function;

[0032] Service coverage sub-objective function for:

[0033] ,

[0034] in, It is the number of patients. Indicates the first The patient reached the first The Euclidean distance to each helicopter deployment location. This refers to the average service radius of the helicopter. For the first The injury severity weighting for each patient;

[0035] Rescue response time sub-objective function for:

[0036] ,

[0037] in, This represents the average flight speed of the helicopter. For the first The distance from each patient to the nearest helicopter deployment point is used to transform the minimization time objective into a maximization problem by employing the reciprocal form;

[0038] Economic cost sub-objective function for:

[0039] ,

[0040] in, It is the fixed cost of a single helicopter and its associated equipment. The variable cost is the fuel cost of each helicopter rescue. Similarly, the cost minimization objective is transformed into a maximization problem by using the reciprocal form.

[0041] Furthermore, in S6, the implementation of the directional learning mechanism in the improved artificial bee colony algorithm is as follows: setting a probability threshold. When generating candidate honey sources, according to Determine whether to search around the current globally optimal honey source;

[0042] The specific formula for the iterative update is as follows:

[0043] ,

[0044] in, This is the updated deployment location. It is the globally optimal deployment position at the current moment. The coordinates of the newly generated candidate honey sources, For the range of values ​​within Random numbers within;

[0045] The final helicopter deployment coordinates are represented as follows .

[0046] The present invention has the following beneficial effects:

[0047] 1) By introducing an injury weighting mechanism and combining it with the weighted K-means algorithm, the initial deployment location of pre-hospital emergency helicopters is determined, so that helicopters can prioritize covering areas where seriously injured patients are concentrated, shorten the rescue response time for critically ill patients, and improve the success rate of treatment.

[0048] 2) Integrate service coverage, response time and economic cost to construct a single-objective fitness function. Through a weight adjustment mechanism, achieve multi-objective synergy within a single optimization framework, taking into account both rescue effectiveness and operational feasibility.

[0049] 3) The weighted K-means algorithm is used to generate the initial deployment locations, providing a high-quality starting point for subsequent optimization. Combined with the elbow method and the contour coefficient method, the optimal number of helicopters to be deployed is objectively determined, avoiding subjective bias from human experience, improving the overall convergence efficiency of the algorithm, and providing data-driven decision-making basis for emergency medical network planning.

[0050] 4) An improved artificial bee colony algorithm is adopted, and a directional learning mechanism is introduced to enhance the ability to finely search the optimal solution region, effectively avoid premature convergence, and improve the solution efficiency and stability.

[0051] 5) The algorithm is based on real data, and the deployment plan is objective and interpretable. It can dynamically adjust the deployment plan according to data from different regions and time periods, and has good scenario adaptability and scalability, making it suitable for the construction of multi-level emergency medical systems.

[0052] 6) The deployment plan for pre-hospital emergency helicopters is automatically generated through algorithms, which improves the utilization efficiency and coverage rationality of helicopters, reduces long-term operating costs, reduces the subjectivity of human planning, and promotes the construction of a smart emergency medical system. Attached Figure Description

[0053] Figure 1 This is a flowchart of the algorithm of the present invention.

[0054] Figure 2 A scatter plot showing the distribution of patients.

[0055] Figure 3 for A line graph of the values ​​and profile coefficient values.

[0056] Figure 4 This is the convergence curve of the fitness function of the IABC algorithm.

[0057] Figure 5 This is a diagram of the final helicopter deployment plan. Detailed Implementation

[0058] The invention will now be further described with reference to the accompanying drawings.

[0059] Figure 1 This is a flowchart of the algorithm of the present invention. The present invention will be further described below with reference to specific embodiments.

[0060] This embodiment uses a simulation experiment as an example. Within a rectangular area with a length of 80,000 meters and a width of 30,000 meters, 2,000 sets of sample points were randomly generated as patient data (including geographical coordinates and injury severity) for this embodiment. The severity of the patients' injuries was divided into five levels, from I to V, with each level accounting for 40%, 30%, 15%, 10%, and 5% respectively. Figure 2 This is a scatter plot showing the distribution of 2000 patients. Furthermore, in this embodiment, the service radius and flight speed of the helicopter are uniform, at 8000 meters and 60 meters per second, respectively. The fixed cost of each helicopter and its supporting facilities is 3 million yuan, and the operating cost, i.e., the fuel consumption for helicopter flight, is 500 yuan per kilometer.

[0061] The specific implementation steps are as follows:

[0062] S1: Geographic coordinates of 2000 patient groups Standardization processes are performed, specifically as follows:

[0063] ,

[0064] ,

[0065] in, Indicates the first The geographical coordinates of each patient This represents the standardized patient location coordinates. and These represent the minimum and maximum values ​​of the geographic coordinates of all patients, respectively.

[0066] S2: Based on the severity of the patient's injury, it can be divided into five injury levels: I, II, III, IV, and V, each assigned a different weight value. (Grade I ~ 1, Grade II ~ 2, Grade III ~ 3, Grade IV ~ 4, Grade V ~ 5), and add this as the third column of data to the patient data to form a weighted dataset. .

[0067] S3: Calculate the weighted cluster squared error and the silhouette coefficient, and determine the optimal number of helicopters to be deployed based on the elbow rule and the silhouette coefficient method as a collaborative decision-making mechanism.

[0068] The elbow rule works by minimizing the sum of squared errors in each cluster subset. The value is used to determine the optimal number of cluster centers. The specific formula changes are as follows:

[0069] ,

[0070] in, This indicates the number of pre-hospital emergency helicopters deployed. , Represents all cluster subsets, For the patient's geographical location, For the first Cluster centers, The patient To the cluster center The square of the distance;

[0071] The silhouette coefficient combines the density and separation of cluster subsets to provide a measurement for each patient. The specific calculation formula is as follows:

[0072] ,

[0073] ,

[0074] in, Indicates the first Profile coefficient of each patient It is the first The average dissimilarity between a patient and other patients within the same cluster subset It is the first Minimum average dissimilarity between a patient and other cluster subsets It is the average of the contour coefficients of all patients.

[0075] By searching By identifying the steepest descent point of the value and the local maximum point of the profile coefficient, the optimal number of pre-hospital emergency helicopters to deploy can be determined. , Figure 3 yes A line graph of the values ​​and profile coefficient values.

[0076] S4: Using a weighted dataset as input, solve for the initial deployment coordinates of k helicopters using a weighted K-means algorithm. Geographic location data for each patient Assign a weight The weight is used to represent the influence of the data on the deployment strategy of pre-hospital emergency helicopters; the higher the weight, the greater the contribution of the data in the clustering process.

[0077] S5: Construct a single-objective fitness function that integrates three sub-functions: service coverage, rescue response time, and economic cost;

[0078] ,

[0079] in, , and These represent the three sub-objective functions: service coverage, rescue response time, and economic cost, respectively. , and This represents the proportion of influence of the three constraints on the objective function (i.e., the weights of the sub-objective functions), which are given based on expert experience and the requirements of the actual scenario. , .

[0080] Service coverage sub-objective function :

[0081] ,

[0082] in, It is the number of patients. It refers to the number of helicopters deployed. It is the first The injury severity weighting for each patient Indicates the first The patient reached the first The Euclidean distance to each helicopter deployment location. It is the average service radius of the helicopter.

[0083] Rescue response time sub-objective function :

[0084] ,

[0085] in, This indicates the average flight speed of the helicopter; It is the first The distance from each patient to the nearest helicopter deployment point is given by the function, which measures the weighted average flight time from the nearest helicopter deployment point to all patients. To transform this into a maximization problem, the reciprocal form is used.

[0086] Economic cost sub-objective function :

[0087] ,

[0088] in, The 10,000 yuan figure represents the fixed cost of a single helicopter and its associated equipment. It is a variable cost, primarily the helicopter's fuel consumption during each rescue operation in this invention. Similarly, it is converted to a reciprocal form.

[0089] S6: By improving the artificial bee colony algorithm to update the deployment location, the final pre-hospital emergency helicopter deployment plan is obtained.

[0090] To accelerate the convergence of the algorithm and utilize the optimal information discovered by the population, this invention introduces a targeted learning mechanism. When generating candidate honey sources, a probability threshold is set... This is used to determine whether a new nectar source is located around the current globally optimal nectar source;

[0091] ,

[0092] in, This is the updated deployment location. It is the globally optimal deployment position at the current moment. These are the coordinates of the newly generated candidate honey sources. It is located in the interval Random numbers within.

[0093] This strategy increases the probability that the population will perform a fine-grained search within the known optimal region, thereby effectively guiding the search direction and accelerating the convergence speed. Figure 4 This is the convergence curve of the fitness function of the IABC algorithm.

[0094] Figure 5 This is the final deployment location for pre-hospital emergency helicopters. .

[0095] The above description is only a preferred embodiment of the present invention. It should be noted that those skilled in the art can make several improvements without departing from the principle of the present invention, and these improvements should also be considered within the scope of protection of the present invention.

Claims

1. A method for optimizing the deployment of pre-hospital emergency helicopters by integrating weighted K-means and IABC, characterized in that: Includes the following steps: S1: Acquire patient data and their geographic coordinates, and perform preprocessing; S2: Based on the severity of the patient's injury, a weighted dataset is formed by classifying the injury and assigning corresponding weights. S3: Calculate the weighted cluster squared error and the silhouette coefficient, and determine the optimal number of helicopters k based on the elbow rule and the silhouette coefficient method as a collaborative decision-making mechanism. S4: Using a weighted dataset as input, solve for the initial deployment coordinates of k helicopters using a weighted K-means algorithm; S5: Construct a single-objective fitness function that integrates three sub-functions: service coverage, rescue response time, and economic cost; S6: An improved artificial bee colony algorithm with an introduced directional learning mechanism is adopted. The initial deployment position of the helicopter is used as the initial population for iterative updates, and the final deployment coordinates of the pre-hospital emergency helicopter are output.

2. The pre-hospital emergency helicopter deployment optimization method integrating weighted K-means and IABC as described in claim 1, characterized in that: In S1, the patient's geographic location coordinates are standardized and preprocessed using the following formula: ， ， in, Indicates the first The geographical coordinates of each patient This represents the standardized geographic coordinates of the patient. and These represent the minimum and maximum values ​​of the geographic coordinates of all patients, respectively.

3. The pre-hospital emergency helicopter deployment optimization method integrating weighted K-means and IABC as described in claim 1, characterized in that: In S2, the weighting of injury classification involves dividing patients into several levels based on the severity of their injury and assigning corresponding weight values ​​to different levels. ; The obtained weight values ​​are used as the third column of data and combined with the standardized geographic coordinates to form a weighted dataset. , This indicates the number of patient data points.

4. The pre-hospital emergency helicopter deployment optimization method integrating weighted K-means and IABC as described in claim 1, characterized in that: In S3, the formula for calculating the weighted cluster squared error sum is: ， in, This indicates the number of pre-hospital emergency helicopters deployed. , Represents all cluster subsets, For the patient's geographical location, For the first Cluster centers, The patient To the cluster center The square of the distance; The formula for calculating the profile coefficient is: ， ， in, Indicates the first Profile coefficient of each patient It is the first The average dissimilarity between a patient and other patients within the same cluster subset It is the first Minimum average dissimilarity between a patient and other cluster subsets It is the average of the contour coefficients of all patients.

5. The pre-hospital emergency helicopter deployment optimization method based on weighted K-means and IABC as described in claim 4, characterized in that: Optimal number of helicopters deployed By searching The steepest point of descent and profile coefficient The local maximum point is determined by collaborative decision-making.

6. The pre-hospital emergency helicopter deployment optimization method integrating weighted K-means and IABC as described in claim 1, characterized in that: In S4, the weighted K-means algorithm uses the weight values ​​in the weighted dataset. Adjust the contribution of each patient's data to the cluster centers, i.e., the weight value. The higher the value, the greater the impact of the corresponding patient data on the clustering process; The initial deployment coordinates of the helicopters are represented as follows: ,Right now The coordinates of the cluster centers.

7. The pre-hospital emergency helicopter deployment optimization method integrating weighted K-means and IABC as described in claim 1, characterized in that: In S5, the specific form of the integrated single-objective fitness function is as follows: ， in, , and These represent the three sub-objective functions: service coverage, rescue response time, and economic cost, respectively. , and The weights represent the proportion of influence of the sub-objective function; Service coverage sub-objective function for: ， in, It is the number of patients. Indicates the first The patient reached the first The Euclidean distance to each helicopter deployment location. This refers to the average service radius of the helicopter. For the first The injury severity weighting for each patient; Rescue response time sub-objective function for: ， in, This represents the average flight speed of the helicopter. For the first The distance from each patient to the nearest helicopter deployment point is used to transform the minimization time objective into a maximization problem by employing the reciprocal form; Economic cost sub-objective function for: ， in, It is the fixed cost of a single helicopter and its associated equipment. The variable cost is the fuel cost of each helicopter rescue. Similarly, the cost minimization objective is transformed into a maximization problem by using the reciprocal form.

8. The pre-hospital emergency helicopter deployment optimization method integrating weighted K-means and IABC as described in claim 1, characterized in that: In S6, the implementation of the directional learning mechanism in the improved artificial bee colony algorithm is as follows: a probability threshold is set. When generating candidate honey sources, according to Determine whether to search around the current globally optimal honey source; The specific formula for the iterative update is as follows: ， in, This is the updated deployment location. It is the globally optimal deployment position at the current moment. The coordinates of the newly generated candidate honey sources, For the range of values ​​within Random numbers within; The final helicopter deployment coordinates are represented as follows .

Images