A dynamic task allocation method integrating task urgency and sensor performance

Abstract

The present invention discloses a dynamic task allocation method that integrates task urgency and sensor performance, including the following steps: Step S1: The task publisher publishes information on the platform; Step S2: After receiving the task information broadcast by the platform, the worker uploads their own information to the platform to compete for the qualification to participate in this task; the platform calculates the balance factor based on the information it has; Step S3: Task allocation, the calculated balance factor is introduced into the allocation algorithm to perform task allocation, and tasks are assigned to workers based on the balance factor. The present invention distinguishes the different sensor performance levels carried by different workers in the mobile crowd-sensing system; and solves the problem that traditional allocation strategies homogenize the performance of device sensors and incorrectly allocate tasks, resulting in poor data quality for high-precision tasks.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a dynamic task allocation method integrating task urgency and sensor performance. Background Art

[0002] Mobile crowdsensing, an emerging data collection and processing model, is widely used in various fields. It can efficiently complete a variety of tasks, such as environmental monitoring, traffic monitoring, and public safety surveillance, leveraging the sensors carried by ordinary users' smart devices. In the field of mobile crowdsensing, the design of task allocation strategies directly determines the efficiency and quality of data collection in complex scenarios. Traditional task allocation methods often optimize based on user location, device availability, or simple reputation models. However, these methods struggle to meet the time-critical or data-quality-sensitive requirements of specific scenarios, such as disaster emergency response, healthcare monitoring, and real-time diagnosis of industrial equipment. Existing technologies suffer from the following significant issues: First, task allocation strategies fail to fully account for the differences in time sensitivity and sensor performance requirements across different tasks. Most existing inventions employ a unified task allocation strategy. This leads to the problem that allocation solutions suitable for time-critical tasks may not be suitable for sensing tasks requiring high data accuracy. Current developments require not only the selection of workers to perform the tasks, but also the matching of these workers with the specific scenario requirements. Second, most current systems assume that all workers' devices have homogeneous sensing capabilities, which is often not the case. With the ever-increasing number and variety of smart devices, standardizing device sensor performance inevitably deviates from current realities and can lead to unnecessary errors. For example, in air quality monitoring, users with low-precision sensors may be misassigned, resulting in inaccurate and unreliable data. Therefore, refined sensor performance is essential for mobile crowdsensing.

[0003] Furthermore, the work habits and skills of the numerous workers in a mobile crowd-sensing system vary widely. For time-sensitive sensing tasks, the platform prioritizes completing them within strict time limits. On most existing platforms, some workers, despite possessing high-performance sensors, may be eliminated from the platform due to their inability to fully meet the time constraints of sensing tasks. This represents a waste of resources for the mobile crowd-sensing system. However, these workers could be retained in the mobile crowd-sensing system by designing a novel task allocation mechanism, leveraging their strengths to perform sensing tasks. A dynamic task allocation mechanism that integrates sensor performance and time constraints could achieve coordinated optimization of sensor performance and task urgency by quantifying the coupling relationship between device sensor capabilities and task urgency constraints. This would allow workers with diverse characteristics to be retained in the same mobile crowd-sensing system, increasing the system's resources and improving task allocation and completion rates. Summary of the Invention

[0004] In view of the shortcomings of the existing technology, the present invention provides a dynamic task allocation method that integrates task urgency and sensor performance, including the following steps:

[0005] Step S1: The task publisher publishes task information on the platform according to the publishing model, and presets the task urgency index and sensor performance index;

[0006] Step S2: Workers receive the task information broadcast by the platform, upload their information to the platform, and compete for the qualification to participate in this task; the balance factor is obtained based on the task urgency index and sensor performance index in step S1;

[0007] Step S3: Task allocation. The task publisher allocates tasks to workers through an allocation algorithm based on the balance factor; the task allocation process is optimized based on the improved genetic algorithm to obtain the optimal task allocation result.

[0008] Furthermore, the publishing model in step S1 is expressed as:

[0009] ;

[0010] ;

[0011] ;

[0012] in, Indicates task information. is the sensor performance index, is the task urgency index, For the task set, Budget for the mission, This is the first task published by the publisher tasks, and They are the first The start and end time of each task, This is the first task published by the publisher The location information of each task, They represent the first The horizontal and vertical coordinates of each task, Indicates the first The sensor type for each task.

[0013] Furthermore, the mathematical model of the worker information in step S2 is as follows:

[0014] ;

[0015] in, For the Worker information, For the Sensor information of each worker, Indicates that the index is The sensor type, Indicates that the index is The sensor performance, Indicates that the index is The sensor type, Indicates that the index is sensor performance; Indicates the The on-time completion status of the most recent task of each worker is 1 or -1, 1 means the task is completed on time; -1 means the task is not completed on time. Indicates the The completion of historical tasks on time; Indicates the The expected bid of each worker for this round of tasks.

[0016] Furthermore, in step S2, workers compete for tasks. At this time, the platform reads the information of each worker participating in the election and comprehensively calculates the equivalent value of the worker's recent task completion on time to obtain the balance factor between the task urgency index and sensor performance.

[0017] The specific process of obtaining the balance factor is as follows: according to the task urgency index set by the task publisher and Sensing Performance Index To determine the time urgency threshold in the allocation algorithm As a balancing factor, it is expressed as:

[0018] ;

[0019] ;

[0020] ;

[0021] ;

[0022] in, Indicates the The equivalent value of the on-time completion of the worker’s most recent task, Indicates the average on-time completion of recent tasks of candidate workers in this round; is the total number of tasks, is the total number of candidate workers, is the influence coefficient; is the efficiency coefficient;

[0023] The task tendency of the task publisher is determined according to the task urgency index and the sensor performance index, and the task is assigned according to the task tendency.

[0024] Furthermore, step S3 includes:

[0025] Step S31: shuffle the order of tasks and assign workers that meet the requirements to each task using an improved genetic algorithm;

[0026] Calculate the work level index of each worker who meets the requirements, sort them, and initially select the worker with the largest work level index value as the worker to perform the task; repeat this process several times to obtain the initial generation population;

[0027] Work Level Index Expressed as:

[0028] ;

[0029] in, For the task The type of sensor required, Represents the performance level equivalent value calculation function.

[0030] Furthermore, step S32: adjusting and replacing workers that do not meet the requirements, checking individuals in the population, and adjusting individuals that do not meet the time urgency threshold; the conditions that need to be met are expressed as follows:

[0031] ;

[0032] in, Indicates a task expenses; Indicates that the worker after replacement Extract the set of sensor types from the information;

[0033] Get a subset of results As an individual in a population, Indicates a task Assigned to workers , Indicates a task Assigned to workers ;

[0034] Step S33: Replace the replaceable workers; the replacement condition is expressed as:

[0035] ;

[0036] in, represents the replaced worker, After replacement budget, Indicates that from the task Extract the required sensor type from Indicates that after the replacement of workers Extract the set of sensor types from the information, Represents the original worker w Required sensor type performance level equivalent, Indicates the replaced worker of Required sensor type performance level equivalence;

[0037] Get a subset of results.

[0038] Furthermore, step S3 further includes:

[0039] Step S34: Calculate population fitness , according to fitness Output the best solution; expressed as:

[0040] ;

[0041] ;

[0042] in, Calculate the function for the equalization factor; is the weight of the mean of the sensor level equivalent value, for The weight of the mean of for the worker's index, is the index of the task, Indicates the total number of tasks, Indicates workers The performance level equivalent value of the sensors required for the corresponding mission carried, Indicates the index range of workers, is the number of groups corresponding to tasks and workers in the individuals of the population;

[0043] After obtaining the fitness values ​​of all individuals in the population, the individual with the maximum fitness is selected as the best allocation scheme.

[0044] Compared with the existing technology, the present invention has the following beneficial effects:

[0045] (1) The present invention improves the task allocation mechanism by introducing a task urgency index and a sensor performance index to integrate sensor performance and task urgency, achieving collaborative optimization. Compared with other inventions, the task allocation mechanism of the present invention can be dynamically adjusted based on the task urgency index and sensor performance index of each round of tasks, and this task allocation mechanism has a wider applicability.

[0046] (2) This invention distinguishes the different sensor performance levels carried by different workers in the mobile crowd-sensing system, solving the problem of traditional allocation strategies that homogenize device sensor performance and incorrectly allocate tasks. It also introduces a balancing factor to achieve a balance between task urgency and sensor performance in the task allocation mechanism. This enables the crowd-sensing platform to maximize the advantages of platform workers according to task requirements.

[0047] (3) Based on the characteristics of the proposed task allocation mechanism, the present invention designs a task allocation algorithm. This algorithm is improved upon the genetic algorithm, and a two-stage optimization scheme is designed to replace the more complex optimization process of the genetic algorithm, which involves crossover and mutation. This algorithm can adapt to the dynamic adjustment characteristics of the task allocation mechanism proposed in the present invention and obtain the optimal task allocation solution. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 This is a system architecture diagram of the present invention.

[0049] Figure 2 It is a comparative experimental diagram of the situation of completing the task on time according to the present invention. DETAILED DESCRIPTION

[0050] refer to Figure 1 , which is the system architecture diagram of the mobile crowd intelligence perception of this application, which is mainly composed of task publishers, mobile crowd intelligence perception platforms and workers who perceive data. The task publisher is the demander and needs to upload the relevant information of the task to the platform. After receiving the information, the platform will broadcast the task to the worker group and recruit workers to perform this task. Then the workers decide whether to participate in this competition. The workers who participate in this competition will form a set of candidate workers. The final task allocation plan is determined by the platform's task matching mechanism to find the best group of workers from the candidate worker set and give the best task allocation plan. After the task allocation is completed, the workers will perform the perception task as agreed and upload the data to the platform. The platform will issue rewards to each worker based on the execution of the task, and then perform some simple processing on the data and send it to the task publisher.

[0051] A dynamic task allocation method integrating task urgency and sensor performance includes the following steps:

[0052] Step S1: The task publisher publishes task information on the platform according to the publishing model, and presets the task urgency index and sensor performance index;

[0053] Step S2: Workers receive the task information broadcast by the platform, upload their information to the platform, and compete for the qualification to participate in this task; the balance factor is obtained based on the task urgency index and sensor performance index in step S1;

[0054] Step S3: Task allocation. The task publisher allocates tasks to workers through an allocation algorithm based on the balance factor; the task allocation process is optimized based on the improved genetic algorithm to obtain the optimal task allocation result.

[0055] Furthermore, the publishing model in step S1 is expressed as:

[0056] ;

[0057] ;

[0058] ;

[0059] in, Indicates task information. is the sensor performance index, is the task urgency index, For the task set, Budget for the mission, This is the first task published by the publisher tasks, and They are the first The start and end time of each task, This is the first task published by the publisher The location information of each task, They represent the first The horizontal and vertical coordinates of each task, Indicates the first The sensor type for each task.

[0060] Furthermore, the mathematical model of the worker information in step S2 is as follows:

[0061] ;

[0062] in, For the Worker information, For the Sensor information of each worker, Indicates that the index is The sensor type, Indicates that the index is The sensor performance, Indicates that the index is The sensor type, Indicates that the index is sensor performance; Indicates the The on-time completion status of the most recent task of each worker is 1 or -1, 1 means the task is completed on time; -1 means the task is not completed on time. Indicates the The completion of historical tasks on time; Indicates the The expected bid of each worker for this round of tasks.

[0063] Furthermore, in step S2, workers compete for tasks. At this time, the platform will read the information of each worker participating in this round of elections and comprehensively calculate the equivalent value of the worker's recent task completion on time to obtain the balance factor between the task urgency index and sensor performance.

[0064] The specific process of obtaining the balance factor is as follows: according to the task urgency index set by the task publisher and Sensing Performance Index To determine the time urgency threshold in the allocation algorithm As a balancing factor, it is expressed as:

[0065] ;

[0066] ;

[0067] ;

[0068] ;

[0069] in, Indicates the The equivalent value of the on-time completion of the worker’s most recent task, Indicates the average on-time completion of recent tasks of candidate workers in this round; is the total number of tasks, is the total number of candidate workers, is the influence coefficient; is the efficiency coefficient;

[0070] The task tendency of the task publisher is determined according to the task urgency index and the sensor performance index, and the task is assigned according to the task tendency.

[0071] Furthermore, step S3 includes:

[0072] Step S31: shuffle the order of tasks and assign workers that meet the requirements to each task using an improved genetic algorithm;

[0073] Calculate the work level index of each worker who meets the requirements, sort them, and initially select the worker with the largest work level index value as the worker to perform the task; repeat this process several times to obtain the initial generation population;

[0074] Work Level Index Expressed as:

[0075] ;

[0076] in, For the task The type of sensor required, Represents the performance level equivalent value calculation function.

[0077] Furthermore, step S32: adjusting and replacing workers that do not meet the requirements, checking individuals in the population, and adjusting individuals that do not meet the time urgency threshold; the conditions that need to be met are expressed as follows:

[0078] ;

[0079] in, Indicates a task expenses; Indicates that the worker after replacement Extract the set of sensor types from the information;

[0080] Get a subset of results As an individual in a population, Indicates a task Assigned to workers , Indicates a task Assigned to workers ;

[0081] Step S33: Replace the replaceable workers; the replacement condition is expressed as:

[0082] ;

[0083] in, represents the replaced worker, After replacement budget, Indicates that from the task Extract the required sensor type from Indicates that after the replacement of workers Extract the set of sensor types from the information, Represents the original worker w Required sensor type performance level equivalent, Indicates the replaced worker of Required sensor type performance level equivalence;

[0084] Get a subset of results.

[0085] Furthermore, step S3 further includes:

[0086] Step S34: Calculate population fitness , according to fitness Output the best solution; expressed as:

[0087] ;

[0088] ;

[0089] in, Calculate the function for the equalization factor; is the weight of the mean of the sensor level equivalent value, for The weight of the mean of for the worker's index, is the index of the task, Indicates the total number of tasks, Indicates workers The performance level equivalent value of the sensors required for the corresponding mission carried, Indicates the index range of workers, is the number of groups corresponding to tasks and workers in the individuals of the population;

[0090] After obtaining the fitness values ​​of all individuals in the population, the individual with the maximum fitness is selected as the best allocation scheme.

[0091] Experimental results

[0092] In step S2, the balance factor is calculated and the threshold of task allocation is calculated based on the balance factor, the time urgency index and the sensor performance index. The result of task allocation will be different depending on the threshold. Figure 2 As shown in the figure, with other factors remaining the same, task assignment is performed using different thresholds. The results show the average on-time completion of the matched workers and their sensor performance. The solid line represents the worker's on-time completion on the right-hand vertical axis, while the dashed line represents the average sensor performance of the workers on the left-hand vertical axis. The horizontal axis represents the threshold value. The figure shows that as the threshold increases, workers' on-time completion improves, but some workers with excellent sensor performance but poor on-time performance are eliminated. The figure also shows that when a task requires as many high-performance sensor workers as possible without a deadline, the threshold should be set lower. Conversely, the threshold can be set higher. This maximizes resource utilization based on the needs of different tasks.

[0093] To further clarify, mobile crowdsensing is a technology that uses smart devices carried by workers to collect data and perform environmental perception. This solution targets two types of workers. The first type of perception workers, such as local residents and the general public, have flexible schedules and are generally able to start and complete perception tasks promptly and on time, strictly adhering to time constraints. However, their sensors are generally of average quality. The second type comprises professionals, such as logistics workers and couriers. These workers are equipped with high-quality GPS modules, RFID readers, and temperature and humidity sensors (for goods requiring environmental control). These sensors are not available to most workers in the first type, and their performance is relatively high. Environmental monitoring tasks requiring high-performance sensors can be assigned to these workers, and their high-quality sensors provide more accurate perception data. Maintenance technicians, who are typically equipped with specialized tools such as voltage detectors and infrared thermal imagers, can help relevant departments obtain timely status information on power lines distributed across different regions, assisting with power maintenance. Vibration sensors and ultrasonic leak detectors can also monitor the usage of gas and water pipes, providing information that can alert relevant departments to perform maintenance. The characteristic of this type of workers is that the sensors they carry are relatively rare and have better performance. However, since they are industry professionals, their time is not as free as that of the first type of workers. This group of people cannot perform strictly time-sensitive sensing tasks. If the platform abandons this type of workers with high-quality sensors for this reason, it will be a huge loss. Therefore, this application calculates the balance factor between the task urgency index and sensor performance of each round based on the task urgency index U and the sensor performance index P, and controls the activation of this type of workers through the time urgency threshold of the task, so that the platform can select workers with higher performance sensors when assigning each round of tasks while meeting the time urgency of the task as much as possible.

[0094] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A dynamic task allocation method integrating task urgency and sensor performance, characterized in that: The steps include: Step S1: The task publisher publishes task information on the platform according to the publishing model, and presets the task urgency index and sensor performance index; Step S2: Workers receive the task information broadcast by the platform, upload their information to the platform, and compete for the qualification to participate in this task; the balance factor is obtained based on the task urgency index and sensor performance index in step S1; Step S3: Task allocation. The task publisher allocates tasks to workers through an allocation algorithm based on the balance factor. The task allocation process is optimized based on an improved genetic algorithm to obtain the optimal task allocation result. The publishing model described in step S1 is expressed as: ； ； ； in, Indicates task information. is the sensor performance index, is the task urgency index, For the task set, Budget for the mission, This is the first task published by the publisher of this task tasks, and They are the first The start and end time of each task, This is the first task published by the publisher of this task The location information of each task, They represent the first The horizontal and vertical coordinates of each task, Indicates the first Sensor type for each task; The mathematical model of the worker information in step S2 is as follows: ； in, For the Worker information, For the Sensor information of each worker, Indicates that the index is The sensor type, Indicates that the index is The sensor performance, Indicates that the index is The sensor type, Indicates that the index is sensor performance; Indicates the The on-time completion status of the most recent task of each worker is 1 or -1, 1 means the task is completed on time; -1 means the task is not completed on time. Indicates the The completion of historical tasks on time; Indicates the The expected bids of workers for this round of tasks; In step S2, workers compete for tasks. At this time, the platform reads the information of each worker participating in the election and comprehensively calculates the equivalent value of the worker's recent task completion on time to obtain the balance factor between the task urgency index and sensor performance. The specific process of obtaining the balance factor is as follows: according to the task urgency index set by the task publisher and Sensing Performance Index To determine the time urgency threshold in the allocation algorithm As a balancing factor, it is expressed as: ； ； ； ； in, Indicates the The equivalent value of the on-time completion of the worker’s most recent task, Indicates the average on-time completion of recent tasks of candidate workers in this round; is the total number of tasks, is the total number of candidate workers, is the influence coefficient; is the efficiency coefficient; Determine the task tendency of the task publisher according to the task urgency index and the sensor performance index, and assign tasks according to the task tendency; Step S3 includes: Step S31: shuffle the order of tasks and assign workers that meet the requirements to each task using a greedy algorithm; Calculate the work level index of each worker who meets the requirements, sort them, and initially select the worker with the largest work level index value as the worker to perform the task; repeat this process several times to obtain the initial generation population; Work Level Index Expressed as: ； in, For the task The type of sensor required, represents the performance level equivalent value calculation function; Step S32: Adjust and replace workers that do not meet the requirements, check the individuals in the population, and adjust the individuals that do not meet the time urgency threshold; the conditions that need to be met are expressed as follows: ； in, Indicates a task expenses; Indicates that the worker after replacement Extract the set of sensor types from the information; Get a subset of results As an individual in a population, Indicates a task Assigned to workers , Indicates a task Assigned to workers ; Step S33: Replace the replaceable workers; the replacement condition is expressed as: ； in, represents the replaced worker, After replacement budget, Indicates that from the task Extract the required sensor type from Indicates that after the replacement of workers Extract the set of sensor types from the information, Represents the original worker w Required sensor type performance level equivalent, Indicates the replaced worker of Required sensor type performance level equivalence; Get a subset of results; Step S3 further includes: Step S34: Calculate population fitness , according to fitness Output the best solution; expressed as: ； ； in, Calculate the function for the equalization factor; is the weight of the mean of the sensor level equivalent value, for The weight of the mean of for the worker's index, is the index of the task, Indicates the total number of tasks, Indicates workers The performance level equivalent value of the sensors required for the corresponding mission carried, Indicates the index range of workers, is the number of groups corresponding to tasks and workers in the individuals of the population; After obtaining the fitness values ​​of all individuals in the population, the individual with the maximum fitness is selected as the best allocation scheme.

Images