Traffic survey system and its resource scheduling method

By using sensors and processors in the traffic survey system to dynamically schedule vehicle information, the problem of unutilized computing resources during traffic troughs is solved, and efficient utilization of resources and cost reduction is achieved.

CN113518110BActive Publication Date: 2024-11-22SHENZHEN INST OF ADVANCED TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202110468575.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-04-28
Publication Date
2024-11-22
Estimated Expiration
2041-04-28

AI Technical Summary

Technical Problem

Existing traffic survey equipment is underutilized during traffic trough periods, resulting in wasted computing resources.

Method used

By introducing servers, sensors, preprocessors and centralized processors into the traffic survey system, using magnetic field signals to collect and process vehicle information, calculate the number of vehicle passages in each lane and the processor utilization rate, realize dynamic resource scheduling, and improve computing resources utilization rate.

Benefits of technology

It effectively improves the utilization rate of computing resources, reduces computing costs, and avoids waste of computing resources while ensuring real-time performance.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN113518110B_ABST
    Figure CN113518110B_ABST
Patent Text Reader

Abstract

The present invention provides a traffic survey system and a resource scheduling method thereof, the resource scheduling method comprising: a sensor collects magnetic field signals of corresponding lanes; a preprocessor obtains signals of vehicles in corresponding lanes according to the magnetic field signals of corresponding lanes; a centralized processor obtains signals of vehicles sent by corresponding preprocessors and respectively calculates the number of vehicles passing through lanes corresponding to the preprocessors within a predetermined time according to the signals of vehicles sent by the preprocessors; the centralized processor calculates the utilization rate of the centralized processor according to the predetermined time, the number of vehicles passing through lanes corresponding to the preprocessors within the predetermined time, and the data processing time of the centralized processor; and a server dynamically schedules multiple centralized processors according to the utilization rates of the multiple centralized processors. The resource scheduling method provided by the present invention dynamically schedules multiple centralized processors according to the utilization rates of the centralized processors, thereby improving the utilization rate of computing resources.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of intelligent transportation technology, and in particular to a traffic survey system and a resource scheduling method thereof. Background Art

[0002] Intelligent transportation needs to process a large number of traffic videos, so it requires a lot of computing resources. The amount of traffic calculation is obviously correlated with the traffic flow. The greater the traffic flow, the more computing resources are needed. The computing power of existing traffic survey equipment is usually designed based on the number of lanes and the number of vehicles during peak traffic hours to meet the real-time computing resources required by traffic survey equipment during busy traffic (rush hour). However, during low traffic hours, the computing resources of computing equipment are not fully utilized, and most of the computing power is idle, which will cause a waste of computing resources. Summary of the invention

[0003] In order to solve the deficiencies of the prior art, the present invention provides a traffic survey system and a resource scheduling method thereof, which performs resource scheduling according to the utilization rate of the centralized processor, effectively improves the utilization rate of computing resources, and reduces computing costs while ensuring real-time performance.

[0004] The specific technical solution proposed by the present invention is: to provide a resource scheduling method for a traffic survey system, the traffic survey system includes a server, multiple sensors, multiple preprocessors and multiple centralized processors, each lane corresponds to two sensors and one preprocessor, each centralized processor corresponds to several preprocessors among the multiple preprocessors, the resource scheduling method includes:

[0005] The sensor collects the magnetic field signal of the corresponding lane;

[0006] Each preprocessor obtains a signal of a vehicle in a corresponding lane according to the magnetic field signal of the corresponding lane;

[0007] Each centralized processor obtains vehicle signals corresponding to a plurality of preprocessors and respectively calculates the number of vehicles passing through lanes corresponding to the plurality of preprocessors within a predetermined time according to the vehicle signals sent by the plurality of preprocessors;

[0008] Each centralized processor calculates the utilization rate of each centralized processor according to a predetermined time, the number of vehicles passing through lanes corresponding to a plurality of pre-processors within the predetermined time, and the data processing time of each centralized processor;

[0009] The server dynamically schedules the multiple centralized processors according to the utilization rates of the multiple centralized processors.

[0010] Furthermore, each preprocessor obtains a signal of a vehicle in a corresponding lane according to the magnetic field signal of the corresponding lane, including:

[0011] Obtain the magnetic field signal of the corresponding lane;

[0012] Sampling the magnetic field signal of the corresponding lane to obtain the voltage signal of the corresponding lane;

[0013] The voltage signal of the corresponding lane is traversed using a sliding window. If the characteristics of the voltage signal in the sliding window meet a predetermined condition, the voltage signal in the sliding window is extracted as a signal of the vehicle in the corresponding lane.

[0014] Further, the characteristics of the voltage signal are the mean square error and the mean voltage of the voltage signal. Accordingly, the predetermined condition is that the mean square error is greater than the mean square error threshold and the mean is greater than the mean threshold.

[0015] Furthermore, the mean square error threshold is 0.005-0.007, the mean value threshold is 0V, and the window width of the sliding window is the width corresponding to a predetermined number of sampling points on the time axis, wherein the predetermined number is 3%-4% of the sensor sampling frequency.

[0016] Furthermore, before each centralized processor calculates the utilization rate of each centralized processor according to the scheduled time, the number of vehicles passing through the lanes corresponding to the plurality of preprocessors within the scheduled time, and the data processing time of each centralized processor, the resource scheduling method further includes:

[0017] Each centralized processor calculates utilization rates of a plurality of preprocessors according to a predetermined time, the number of vehicles passing through lanes corresponding to the plurality of preprocessors within the predetermined time, and data processing time of the plurality of preprocessors;

[0018] Determining whether there is a preprocessor whose utilization rate is less than the second maximum utilization rate among the plurality of preprocessors;

[0019] Accordingly, each centralized processor calculates the utilization rate of each centralized processor according to the scheduled time, the number of vehicles passing through the lanes corresponding to the plurality of preprocessors within the scheduled time, and the data processing time of each centralized processor, including:

[0020] If there is a preprocessor among several preprocessors whose utilization rate is less than the second maximum utilization rate, each centralized processor calculates the utilization rate of each centralized processor based on the scheduled time, the number of vehicles passing through the lanes corresponding to the several preprocessors within the scheduled time, and the data processing time of each centralized processor.

[0021] Furthermore, the calculation formula for the utilization of each preprocessor is:

[0022]

[0023] Among them, Upre represents the utilization rate of each preprocessor, k represents the number of vehicles passing through the lane corresponding to each preprocessor within the predetermined time T, and t p It represents the time required for each preprocessor to process the data of a vehicle, t s It represents the time required for each preprocessor to transmit data to the corresponding centralized processor, a represents the A / D resolution of each preprocessor, b represents the data cache size of each preprocessor, and l represents the number of sampling points for each vehicle.

[0024] Furthermore, the calculation formula for the utilization of each centralized processor is:

[0025]

[0026] Among them, U ctr represents the utilization rate of each centralized processor, k represents the number of vehicles passing through the lane corresponding to each preprocessor within the predetermined time T, and t s It represents the time required for each preprocessor to transmit data to the corresponding central processor, t c represents the time required for each centralized processor to process the data of a vehicle, a represents the A / D resolution of each preprocessor, b represents the data cache size of each preprocessor, l represents the number of sampling points for each vehicle, and t syn It represents the time required for each centralized processor to send data to the server, and n represents the number of pre-processors corresponding to each centralized processor.

[0027] Furthermore, the server dynamically schedules the multiple centralized processors according to the utilization rates of the multiple centralized processors, including:

[0028] Determining whether the utilization rate of each of the plurality of centralized processors is less than a first maximum utilization rate;

[0029] If the utilization rate of the centralized processor is less than the first maximum utilization rate, the centralized processor is used as a centralized processor to be scheduled, and a plurality of centralized processors to be scheduled are obtained;

[0030] The multiple centralized processors to be scheduled are sorted according to the size of the utilization rate to obtain a sequence of centralized processors to be scheduled, and the following scheduling process is performed on the sequence of centralized processors to be scheduled:

[0031] The centralized processor to be scheduled with the lowest utilization in the centralized processor sequence to be scheduled is used as the scheduling centralized processor, and the centralized processor to be scheduled with the highest utilization in the centralized processor sequence to be scheduled is used as the merged centralized processor, and it is determined whether the sum of the utilizations of the scheduling centralized processor and the merged centralized processor is less than a first maximum utilization;

[0032] If the sum of the utilization rates of the scheduling centralized processor and the merged centralized processor is less than the first maximum utilization rate, the excess computing resources of the merged centralized processor are allocated to the scheduling centralized processor and the scheduling centralized processor and the merged centralized processor are deleted from the sequence of centralized processors to be scheduled; if the sum of the utilization rates of the scheduling centralized processor and the merged centralized processor is greater than the first maximum utilization rate, the merged centralized processor is deleted from the sequence of centralized processors to be scheduled, and the scheduling process is executed in a loop.

[0033] The present invention also provides a traffic survey system, which includes a server, multiple sensors, multiple preprocessors, and multiple centralized processors, each lane corresponds to two sensors and one preprocessor, and each centralized processor corresponds to several preprocessors among the multiple preprocessors;

[0034] Each sensor is used to collect the magnetic field signal of the corresponding lane within a predetermined time;

[0035] Each preprocessor is used to obtain a signal of a vehicle passing through the corresponding lane within a predetermined time according to the magnetic field signal of the corresponding lane within a predetermined time;

[0036] Each centralized processor is used to obtain vehicle signals sent by the corresponding preprocessors and calculate the number of vehicles passing through the lanes corresponding to the preprocessors within a predetermined time according to the vehicle signals sent by the preprocessors, and to calculate the utilization rate of each centralized processor according to the predetermined time, the number of vehicles passing through the lanes corresponding to the preprocessors within the predetermined time, and the data processing time of each centralized processor;

[0037] The server is used to dynamically schedule multiple centralized processors according to utilization rates of the multiple centralized processors.

[0038] Further, each preprocessor includes an A / D sampling module and a vehicle signal extraction module;

[0039] An A / D sampling module is used to sample the magnetic field signal of the corresponding lane within a predetermined time to obtain the voltage signal of the corresponding lane within the predetermined time;

[0040] The vehicle signal extraction module is used to traverse the voltage signal using a sliding window, and if the characteristics of the voltage signal in the sliding window meet a predetermined condition, the voltage signal in the sliding window is extracted as a vehicle signal.

[0041] The resource scheduling method of the traffic survey system provided by the present invention is that after the sensor obtains the magnetic field signal of the corresponding lane, the preprocessor extracts the signal of the vehicle in the corresponding lane from it, and the centralized processor calculates the number of vehicles passing through the lanes corresponding to the several preprocessors within a predetermined time according to the vehicle signals sent by the several preprocessors, and calculates the utilization rate of each centralized processor according to the predetermined time, the number of vehicles passing through the lanes corresponding to the several preprocessors within the predetermined time and the data processing time of each centralized processor. Finally, the server dynamically schedules multiple centralized processors according to the utilization rate of each centralized processor, thereby realizing the value of mining idle computing resources and improving the utilization rate of computing resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The technical solutions and other beneficial effects of the present invention will be made apparent by describing in detail the specific embodiments of the present invention in conjunction with the accompanying drawings.

[0043] Figure 1 A schematic diagram of the structure of a traffic survey system in an embodiment of the present invention;

[0044] Figure 2 A schematic diagram of a resource scheduling method according to an embodiment of the present invention;

[0045] Figure 3 is a schematic diagram of a sliding window in an embodiment of the present invention;

[0046] Figure 4 A schematic diagram of the connection between a server and a plurality of centralized processors and between each centralized processor and a plurality of pre-processors in an embodiment of the present invention;

[0047] Figure 5 It is another connection diagram between a server and a plurality of central processors and between each central processor and a plurality of pre-processors in an embodiment of the present invention. DETAILED DESCRIPTION

[0048] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. However, the present invention may be implemented in many different forms, and the present invention should not be construed as being limited to the specific embodiments set forth herein. On the contrary, these embodiments are provided to explain the principles of the present invention and their practical applications, so that other persons skilled in the art can understand the various embodiments of the present invention and the various modifications suitable for specific intended applications. In the accompanying drawings, the same reference numerals will always be used to represent the same elements.

[0049] Reference Figure 1The traffic survey system in this embodiment includes a server, multiple sensors, multiple preprocessors, and multiple centralized processors. Each lane corresponds to two sensors and one preprocessor, and each centralized processor corresponds to several preprocessors in the multiple preprocessors. The server, multiple preprocessors, and multiple centralized processors are integrated into one device and installed beside the road.

[0050] Reference Figure 2 The resource scheduling method of the traffic survey system in this embodiment includes the following steps:

[0051] S1, the sensor collects the magnetic field signal of the corresponding lane;

[0052] S2, each preprocessor obtains a signal of a vehicle in a corresponding lane according to the magnetic field signal of the corresponding lane;

[0053] S3, each centralized processor obtains vehicle signals sent by corresponding preprocessors and calculates the number of vehicles passing through lanes corresponding to the preprocessors within a predetermined time according to the vehicle signals sent by the preprocessors;

[0054] S4, each centralized processor calculates the utilization rate of each centralized processor according to the scheduled time, the number of vehicles passing through the lanes corresponding to the plurality of pre-processors within the scheduled time, and the data processing time of each centralized processor;

[0055] S5. The server dynamically schedules the multiple centralized processors according to the utilization rates of the multiple centralized processors.

[0056] In step S1, the sensor in this embodiment is a magnetic field sensor. Two sensors in each lane are buried in the middle of the lane. When a vehicle in the lane passes, a magnetic field is generated. The two sensors are used to obtain the magnetic field signal generated when the vehicle passes. Preferably, the distance between the two sensors is 1 meter, and the sampling frequency of each sensor is 1000Hz. Among them, the two sensors collect the magnetic field signal of the corresponding lane in real time. The two sensors in each lane are connected to the preprocessor corresponding to the lane through a communication cable to send the received magnetic field signal to the preprocessor.

[0057] In step S2, since the magnetic field signal sent by the sensor is an analog signal, and the data type processed by the preprocessor is a digital signal, the preprocessor needs to convert the magnetic field signal into a digital signal after receiving the magnetic field signal of the sensor, and then the preprocessor processes the digital signal to obtain the signal of the vehicle in the corresponding lane. Specifically, step S2 includes:

[0058] S21, obtaining a magnetic field signal of a corresponding lane;

[0059] S22, sampling the magnetic field signal of the corresponding lane to obtain a voltage signal of the corresponding lane;

[0060] S23, using a sliding window to traverse the voltage signal of the corresponding lane, if the characteristics of the voltage signal in the sliding window meet a predetermined condition, extracting the voltage signal in the sliding window as the signal of the vehicle in the corresponding lane.

[0061] In step S22, the preprocessor performs analog-to-digital conversion by sampling the magnetic field signal of the corresponding lane to obtain the voltage signal of the corresponding lane, wherein the voltage signal is a digital signal, and the voltage signal also includes the vehicle signal and other environmental signals. In step S23, this embodiment extracts the vehicle signal by adopting a sliding window traversal method. When there is no vehicle passing through the lane, the characteristics of the voltage signal sampled by the preprocessor are relatively stable. When there is a vehicle passing through the lane, the characteristics of the voltage signal sampled by the preprocessor fluctuate greatly. Therefore, the vehicle signal can be extracted by judging whether the characteristics of the voltage signal in the sliding window meet the predetermined conditions. Specifically, if Figure 3 As shown, the voltage signal is traversed by sliding a sliding window along the time axis. If the characteristics of the voltage signal in the sliding window meet the predetermined conditions, the voltage signal in the sliding window is extracted as the vehicle signal and the vehicle signal is cached.

[0062] Specifically, the characteristics of the voltage signal in this embodiment refer to the mean square error and the mean voltage of the voltage signal. Correspondingly, the predetermined condition refers to that the mean square error is greater than the mean square error threshold and the mean is greater than the mean threshold, that is, if the mean square error of the voltage signal in the sliding window is greater than the mean square error threshold and the mean of the voltage signal in the sliding window is greater than the mean threshold, then the voltage signal in the sliding window is a vehicle signal, and the signal is extracted and cached.

[0063] The mean square error threshold in this embodiment is 0.005-0.007, and the mean threshold is 0V, wherein the mean threshold is set according to the reference voltage sampled by the preprocessor. Preferably, the mean threshold is equal to the reference voltage sampled by the preprocessor, and the mean square error threshold is 0.006. The window width of the sliding window is the width corresponding to a predetermined number of sampling points on the time axis, wherein the predetermined number is 3%-4% of the sampling frequency of the sensor. For example, if the sampling frequency of the sensor is 1000Hz, the window width of the sliding window is the width corresponding to 30-40 sampling points on the time axis, that is, the sliding window includes 30-40 sampling points. Preferably, the window width of the sliding window is the width corresponding to 30 sampling points on the time axis.

[0064] In step S3, the preprocessor sends the cached vehicle signals to the centralized processor, and the centralized processor uses the highest priority method to check whether there are cached vehicle signals in the corresponding preprocessors. If there are, the cached vehicle signals are captured immediately. The centralized processor extracts the number of vehicle signals of the preprocessors within a predetermined time from the vehicle signals sent by the preprocessors, and then obtains the number of vehicles passing through the lanes corresponding to the preprocessors within the predetermined time based on the number of vehicle signals of the preprocessors within the predetermined time. Specifically, the number of vehicles passing through the lane corresponding to each preprocessor within the predetermined time is equal to the number of vehicle signals of each preprocessor within the predetermined time.

[0065] In step S4, each centralized processor calculates the utilization rate of each centralized processor according to the predetermined time, the number of vehicles passing through the lanes corresponding to the plurality of preprocessors within the predetermined time, and the data processing time of each centralized processor using the following formula:

[0066]

[0067] Among them, U ctr represents the utilization rate of each centralized processor, k represents the number of vehicles passing through the lane corresponding to each preprocessor within the predetermined time T, and t s It represents the time required for each preprocessor to transmit data to the corresponding central processor, t c represents the time required for each centralized processor to process the data of a vehicle, a represents the A / D resolution of each preprocessor, b represents the data cache size of each preprocessor, l represents the number of sampling points for each vehicle, and t syn represents the time required for each centralized processor to send data to the server, and n represents the number of preprocessors corresponding to each centralized processor. It should be noted here that the frequency of vehicles passing through each lane within the predetermined time T remains unchanged, and the time required for each centralized processor to process the data of a vehicle is the time required to calculate various characteristics of the vehicle. In this embodiment, the time required to calculate various characteristics of the vehicle includes but is not limited to the time required to calculate the length, speed, category, etc. of the vehicle. The time required for each centralized processor to send data to the server includes the time to package according to the data format of the server and the time to send the packaged data to the server.

[0068] In step S5, Figures 4-5 As shown, Figure 4 The figure shows the connection relationship between the pre-scheduling server and multiple centralized processors and between each centralized processor and several pre-processors. Figure 5The connection relationship between the server and multiple centralized processors and each centralized processor and several preprocessors after scheduling is shown. The server dynamically schedules multiple centralized processors according to the utilization rate of each centralized processor. Specifically, for a centralized processor with a low utilization rate, the server can process the data service corresponding to the centralized processor by scheduling the computing resources of other centralized processors, and then schedule the idle computing resources of the centralized processor to other tasks, thereby realizing the value of mining idle computing resources and improving the utilization rate of computing resources.

[0069] Specifically, the server dynamically schedules the multiple centralized processors according to the utilization rates of the multiple centralized processors respectively, including:

[0070] S51, determining whether the utilization rate of each of the plurality of centralized processors is less than a first maximum utilization rate, if the utilization rate of the centralized processor is less than the first maximum utilization rate, taking the centralized processor as a centralized processor to be scheduled, and proceeding to step S52;

[0071] S52, sorting the centralized processors to be scheduled according to their utilization rates, obtaining a sequence of centralized processors to be scheduled, and performing the following scheduling process on the sequence of centralized processors to be scheduled;

[0072] The centralized processor to be scheduled with the lowest utilization in the centralized processor sequence to be scheduled is used as the scheduling centralized processor, and the centralized processor to be scheduled with the highest utilization in the centralized processor sequence to be scheduled is used as the merged centralized processor, and it is determined whether the sum of the utilizations of the scheduling centralized processor and the merged centralized processor is less than the first maximum utilization;

[0073] If the sum of the utilization rates of the scheduling centralized processor and the merged centralized processor is less than the first maximum utilization rate, the excess computing resources of the merged centralized processor are allocated to the scheduling centralized processor and the scheduling centralized processor and the merged centralized processor are deleted from the sequence of centralized processors to be scheduled; if the sum of the utilization rates of the scheduling centralized processor and the merged centralized processor is greater than the first maximum utilization rate, the merged centralized processor is deleted from the sequence of centralized processors to be scheduled, and the above scheduling process is executed repeatedly until there is no scheduling centralized processor that meets the conditions in the sequence of centralized processors to be scheduled.

[0074] In step S51, the first maximum utilization rate refers to the utilization rate corresponding to the centralized processor working at full load. The above scheduling process is described below through a specific example. Assume that after step S51, J centralized processors to be scheduled are obtained, and the utilization rates of these J centralized processors to be scheduled are all less than the first maximum utilization rate. Each centralized processor to be scheduled is Sort the J centralized processors to be scheduled in ascending order of utilization to obtain a sequence of centralized processors to be scheduled. and maximum The utilization is merged. If the combined utilization is less than the first maximum utilization, the resources of the two centralized processors are scheduled. Specifically, the maximum utilization The corresponding centralized processor's excess computing resources are allocated to the minimum The corresponding centralized processor will minimize The corresponding centralized processors are assigned to other tasks. The centralized processors that are successfully scheduled are deleted from the centralized processor sequence to be scheduled, and the remaining centralized processors in the current centralized processor sequence to be scheduled are used as the centralized processor sequence to be scheduled in the next round of scheduling, and the next round of scheduling is merged. If the utilization rate after the merger is not less than the first maximum utilization rate, that is, in is the first maximum utilization, then the maximum The corresponding centralized processor is deleted from the centralized processor sequence to be scheduled, and the remaining centralized processors to be scheduled in the current centralized processor sequence to be scheduled are used as the centralized processor sequence to be scheduled for the next round of scheduling, and the next round of scheduling merging is performed until there are no scheduling centralized processors that meet the conditions in the centralized processor sequence to be scheduled.

[0075] In this embodiment, the server performs dynamic scheduling in real time. As long as the dynamic scheduling conditions are met, the above scheduling process is executed. For the merged centralized processor that allocates computing resources, if the utilization rate of the merged centralized processor subsequently reaches the first maximum utilization rate, the server will reschedule the computing resources allocated by the merged centralized processor.

[0076] The dynamic scheduling in this embodiment is performed only when there are available computing resources in the centralized processor of the entire traffic survey system. Whether there are available computing resources in the centralized processor is determined based on the utilization of the corresponding preprocessors. If the corresponding preprocessors are all working at full capacity, the centralized processor is also working at full capacity, that is, the centralized processor has no available computing resources. Only when there is a preprocessor that is not working at full capacity among the corresponding preprocessors, the centralized processor will have available resources. Therefore, before calculating the utilization of the centralized processor, it will be preliminarily determined whether the centralized processor is available based on the utilization of the preprocessor, that is, before step S4, the resource scheduling method in this embodiment also includes:

[0077] S40, each centralized processor calculates utilization rates of a plurality of preprocessors according to a predetermined time, the number of vehicles passing through lanes corresponding to the plurality of preprocessors within the predetermined time, and data processing time of the plurality of preprocessors;

[0078] S41. Determine whether there is a preprocessor whose utilization rate is less than the second maximum utilization rate among the plurality of preprocessors. If so, proceed to step S4.

[0079] In step S40, the utilization rate of each preprocessor is calculated by the following formula:

[0080]

[0081] Among them, U pre represents the utilization rate of each preprocessor, k represents the number of vehicles passing through the lane corresponding to each preprocessor within the predetermined time T, and t p It represents the time required for each preprocessor to process the data of a vehicle, t s It indicates the time required for each preprocessor to transmit data to the corresponding centralized processor, a indicates the A / D resolution of each preprocessor, b indicates the data cache of each preprocessor, and l indicates the number of sampling points of each vehicle. It should be noted that the time required for each preprocessor to process the data of a vehicle includes the time for sampling the magnetic field signal of the corresponding lane and extracting the vehicle signal.

[0082] In step S41, the second maximum utilization refers to the utilization corresponding to the preprocessor working at full load. The computing resources of the centralized processor will be scheduled only when there are preprocessors whose utilizations are all less than the second maximum utilization among the preprocessors corresponding to the centralized processor.

[0083] The traffic investigation system in this embodiment is described in detail below. The server is an intelligent traffic system server, for example, Intelligent Traffic System (ITS) or Intelligent Traffic System (ITS) server. The preprocessor may be, for example, an ARM Cortex M0+ chip, and the centralized processor may be, for example, a Linux operating system or an x86 processor. Of course, the preprocessor and the centralized processor may also be other types of processors, which are not limited here.

[0084] Specifically, each sensor is used to collect the magnetic field signal of the corresponding lane within a predetermined time, each preprocessor is used to obtain the signal of the vehicle passing through the corresponding lane within the predetermined time according to the magnetic field signal of the corresponding lane within the predetermined time, each centralized processor is used to obtain the signal of the vehicle sent by the corresponding several preprocessors and calculate the number of vehicles passing through the lanes corresponding to the several preprocessors within the predetermined time according to the signals of the vehicles sent by the several preprocessors, and is used to calculate the utilization rate of each centralized processor according to the predetermined time, the number of vehicles passing through the lanes corresponding to the several preprocessors within the predetermined time and the data processing time of each centralized processor, and the server is used to dynamically schedule multiple centralized processors according to the utilization rates of the multiple centralized processors.

[0085] Each preprocessor includes an A / D sampling module and a vehicle signal extraction module. The A / D sampling module is used to sample the magnetic field signal of the corresponding lane within a predetermined time to obtain the voltage signal of the corresponding lane within the predetermined time. The vehicle signal extraction module is used to traverse the voltage signal of the corresponding lane using a sliding window. If the characteristics of the voltage signal within the sliding window meet the predetermined conditions, the voltage signal within the sliding window is extracted as the signal of the vehicle in the corresponding lane.

[0086] The centralized processor in this embodiment is also used to calculate the utilization rates of several preprocessors according to the scheduled time, the number of vehicles passing through the lanes corresponding to the several preprocessors within the scheduled time, and the data processing time of the several preprocessors, and to determine whether there is a preprocessor among the several preprocessors whose utilization rate is less than the second maximum utilization rate. If there is a preprocessor among the several preprocessors whose utilization rate is less than the second maximum utilization rate, then each centralized processor calculates the utilization rate of each centralized processor according to the scheduled time, the number of vehicles passing through the lanes corresponding to the several preprocessors within the scheduled time, and the data processing time of each centralized processor.

[0087] The above is only a specific implementation of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A resource scheduling method for a traffic survey system, characterized in that: The traffic survey system includes a server, multiple sensors, multiple preprocessors and multiple centralized processors, each lane corresponds to two sensors and one preprocessor, each centralized processor corresponds to several preprocessors among the multiple preprocessors, and the resource scheduling method includes: The sensor collects the magnetic field signal of the corresponding lane; Each preprocessor obtains a signal of a vehicle in a corresponding lane according to the magnetic field signal of the corresponding lane; Each centralized processor obtains vehicle signals corresponding to a plurality of preprocessors and respectively calculates the number of vehicles passing through lanes corresponding to the plurality of preprocessors within a predetermined time according to the vehicle signals sent by the plurality of preprocessors; Each centralized processor calculates the utilization rate of each centralized processor according to a predetermined time, the number of vehicles passing through lanes corresponding to a plurality of pre-processors within the predetermined time, and the data processing time of each centralized processor; The server dynamically schedules the multiple centralized processors according to the utilization rates of the multiple centralized processors; Before each centralized processor calculates the utilization rate of each centralized processor according to the scheduled time, the number of vehicles passing through the lanes corresponding to the plurality of preprocessors within the scheduled time, and the data processing time of each centralized processor, the resource scheduling method further includes: Each centralized processor calculates utilization rates of a plurality of preprocessors according to a predetermined time, the number of vehicles passing through lanes corresponding to the plurality of preprocessors within the predetermined time, and data processing time of the plurality of preprocessors; Determining whether there is a preprocessor whose utilization rate is less than the second maximum utilization rate among the plurality of preprocessors; Accordingly, each centralized processor calculates the utilization rate of each centralized processor according to the scheduled time, the number of vehicles passing through the lanes corresponding to the plurality of preprocessors within the scheduled time, and the data processing time of each centralized processor, including: If there is a preprocessor with a utilization rate less than the second maximum utilization rate among the plurality of preprocessors, each centralized processor calculates the utilization rate of each centralized processor according to the predetermined time, the number of vehicles passing through the lanes corresponding to the plurality of preprocessors within the predetermined time, and the data processing time of each centralized processor; The calculation formula of the utilization rate of each preprocessor is: Among them, U pre represents the utilization rate of each preprocessor, k represents the number of vehicles passing through the lane corresponding to each preprocessor within the predetermined time T, and t p It represents the time required for each preprocessor to process the data of a vehicle, t s It represents the time required for each preprocessor to transmit data to the corresponding centralized processor, a represents the A / D resolution of each preprocessor, b represents the data cache size of each preprocessor, and l represents the number of sampling points for each vehicle.

2. The resource scheduling method according to claim 1, characterized in that: Each preprocessor obtains a signal of a vehicle in a corresponding lane according to the magnetic field signal of the corresponding lane, including: Obtain the magnetic field signal of the corresponding lane; Sampling the magnetic field signal of the corresponding lane to obtain the voltage signal of the corresponding lane; The voltage signal of the corresponding lane is traversed using a sliding window. If the characteristics of the voltage signal in the sliding window meet a predetermined condition, the voltage signal in the sliding window is extracted as a signal of the vehicle in the corresponding lane.

3. The resource scheduling method according to claim 2, characterized in that: The characteristics of the voltage signal are the mean square error and the mean voltage of the voltage signal. Accordingly, the predetermined condition is that the mean square error is greater than the mean square error threshold and the mean is greater than the mean threshold.

4. The resource scheduling method according to claim 3, characterized in that: The mean square error threshold is 0.005-0.007, the mean value threshold is 0V, and the window width of the sliding window is the width corresponding to a predetermined number of sampling points on the time axis, wherein the predetermined number is 3%-4% of the sensor sampling frequency.

5. The resource scheduling method according to claim 1, characterized in that: The calculation formula for the utilization of each centralized processor is: Among them, U ctr represents the utilization rate of each centralized processor, k represents the number of vehicles passing through the lane corresponding to each preprocessor within the predetermined time T, and t s It represents the time required for each preprocessor to transmit data to the corresponding central processor, t c represents the time required for each centralized processor to process the data of a vehicle, a represents the A / D resolution of each preprocessor, b represents the data cache size of each preprocessor, l represents the number of sampling points of each vehicle, and t syn It represents the time required for each centralized processor to send data to the server, and n represents the number of pre-processors corresponding to each centralized processor.

6. The resource scheduling method according to claim 1, characterized in that: The server dynamically schedules the multiple centralized processors according to the utilization rates of the multiple centralized processors, including: Determining whether the utilization rate of each of the plurality of centralized processors is less than a first maximum utilization rate; If the utilization rate of the centralized processor is less than the first maximum utilization rate, the centralized processor is used as a centralized processor to be scheduled, and a plurality of centralized processors to be scheduled are obtained; The multiple centralized processors to be scheduled are sorted according to the size of the utilization rate to obtain a sequence of centralized processors to be scheduled, and the following scheduling process is performed on the sequence of centralized processors to be scheduled: The centralized processor to be scheduled with the lowest utilization in the centralized processor sequence to be scheduled is used as the scheduling centralized processor, and the centralized processor to be scheduled with the highest utilization in the centralized processor sequence to be scheduled is used as the merged centralized processor, and it is determined whether the sum of the utilizations of the scheduling centralized processor and the merged centralized processor is less than a first maximum utilization; If the sum of the utilization rates of the scheduling centralized processor and the merged centralized processor is less than the first maximum utilization rate, the excess computing resources of the merged centralized processor are allocated to the scheduling centralized processor and the scheduling centralized processor and the merged centralized processor are deleted from the sequence of centralized processors to be scheduled; if the sum of the utilization rates of the scheduling centralized processor and the merged centralized processor is greater than the first maximum utilization rate, the merged centralized processor is deleted from the sequence of centralized processors to be scheduled, and the scheduling process is executed in a loop.

7. A traffic survey system, characterized in that: The traffic survey system includes a server, a plurality of sensors, a plurality of preprocessors and a plurality of centralized processors, each lane corresponds to two sensors and a preprocessor, and each centralized processor corresponds to a plurality of preprocessors among the plurality of preprocessors; Each sensor is used to collect the magnetic field signal of the corresponding lane within a predetermined time; Each preprocessor is used to obtain a signal of a vehicle passing through the corresponding lane within a predetermined time according to the magnetic field signal of the corresponding lane within a predetermined time; Each centralized processor is used to obtain vehicle signals sent by the corresponding preprocessors and calculate the number of vehicles passing through the lanes corresponding to the preprocessors within a predetermined time according to the vehicle signals sent by the preprocessors, and to calculate the utilization rate of each centralized processor according to the predetermined time, the number of vehicles passing through the lanes corresponding to the preprocessors within the predetermined time, and the data processing time of each centralized processor; The server is used to dynamically schedule the multiple centralized processors according to the utilization rates of the multiple centralized processors respectively; Before each centralized processor calculates the utilization rate of each centralized processor according to the predetermined time, the number of vehicles passing through the lanes corresponding to the plurality of preprocessors within the predetermined time, and the data processing time of each centralized processor, each centralized processor is further used to: Calculating utilization rates of the plurality of preprocessors according to a predetermined time, the number of vehicles passing through lanes corresponding to the plurality of preprocessors within the predetermined time, and data processing time of the plurality of preprocessors; Determining whether there is a preprocessor whose utilization rate is less than the second maximum utilization rate among the plurality of preprocessors; Accordingly, each centralized processor calculates the utilization rate of each centralized processor according to the scheduled time, the number of vehicles passing through the lanes corresponding to the plurality of preprocessors within the scheduled time, and the data processing time of each centralized processor, including: If there is a preprocessor with a utilization rate less than the second maximum utilization rate among the plurality of preprocessors, each centralized processor calculates the utilization rate of each centralized processor according to the predetermined time, the number of vehicles passing through the lanes corresponding to the plurality of preprocessors within the predetermined time, and the data processing time of each centralized processor; The calculation formula of the utilization rate of each preprocessor is: Among them, U pre represents the utilization rate of each preprocessor, k represents the number of vehicles passing through the lane corresponding to each preprocessor within the predetermined time T, and t p It represents the time required for each preprocessor to process the data of a vehicle, t s It represents the time required for each preprocessor to transmit data to the corresponding centralized processor, a represents the A / D resolution of each preprocessor, b represents the data cache size of each preprocessor, and l represents the number of sampling points for each vehicle.

8. The traffic survey system according to claim 7, characterized in that: Each preprocessor includes an A / D sampling module and a vehicle signal extraction module; An A / D sampling module is used to sample the magnetic field signal of the corresponding lane within a predetermined time to obtain the voltage signal of the corresponding lane within the predetermined time; The vehicle signal extraction module is used to traverse the voltage signal using a sliding window, and if the characteristics of the voltage signal in the sliding window meet a predetermined condition, the voltage signal in the sliding window is extracted as a vehicle signal.