Network dynamic optimization method and system for computer architecture

Abstract

The invention relates to the field of network optimization, in particular to a network dynamic optimization method and system for computer architecture. A network dynamic optimization system of a computer architecture comprises an environment evaluation module, a task weight evaluation module and a resource allocation module. According to the method, the network state, the vehicle speed and the road congestion degree of the unmanned vehicle are monitored in real time, allocation of computing resources and network resources can be dynamically adjusted, and a self-adaptive adjustment mechanism of the method enables the unmanned vehicle to timely optimize a resource allocation strategy according to environment data and task requirements which change in real time; in the face of rapidly changing network conditions and traffic conditions, the excellent adaptability and real-time performance of the method can ensure that the driving performance and the decision-making efficiency of the unmanned vehicle are always kept in the optimal state, so that the response speed and the adaptive capacity of the vehicle to a complex traffic environment are improved.

Description

Technical Field

[0001] The present invention relates to the field of network optimization, and in particular to a method and system for dynamic network optimization of a computer architecture. Background Art

[0002] With the rapid development of computer technology and network technology, computer systems and network environments have become an indispensable infrastructure in modern society. Especially in the field of unmanned driving, dynamic network optimization of computer architecture has become particularly important. Unmanned vehicles rely on powerful computing power and efficient network communication to ensure driving safety and real-time decision-making. In existing technologies, network optimization methods mainly focus on static network configuration and bandwidth management. These methods fail to fully consider factors such as the complex and changeable network status, vehicle speed changes, and road congestion encountered by unmanned vehicles during actual driving. In addition, traditional optimization algorithms often adopt fixed resource allocation strategies and lack adaptability and dynamic adjustment capabilities to real-time environmental data, which makes them incapable of coping with rapidly changing network conditions and traffic conditions.

[0003] In order to solve the above problems, a network dynamic optimization method and system of computer architecture are proposed. The method aims to optimize the driving performance and decision-making efficiency of unmanned vehicles by dynamically adjusting the allocation of computing resources and network resources through real-time monitoring of the network status, vehicle speed and road congestion of unmanned vehicles. It can adaptively adjust the resource allocation strategy according to real-time environmental data and task requirements, so that unmanned vehicles can achieve more efficient and intelligent driving decisions while ensuring safety. Summary of the invention

[0004] The present invention can dynamically adjust the allocation of computing resources and network resources by real-time monitoring of the network status, vehicle speed and road congestion of the unmanned vehicle. The adaptive adjustment mechanism of this method enables the unmanned vehicle to instantly optimize the resource allocation strategy according to the real-time changing environmental data and task requirements; in the face of rapidly changing network conditions and traffic conditions, its excellent adaptability and real-time performance can ensure that the driving performance and decision-making efficiency of the unmanned vehicle are always maintained in the best state, thereby improving the vehicle's response speed and adaptability to complex traffic environments.

[0005] A method for dynamic network optimization of a computer architecture, comprising:

[0006] At any monitoring time point during the driving process of the unmanned vehicle, the vehicle's current environmental data is obtained, including network status data, vehicle speed data, and observed road section vehicle data; based on the obtained environmental data, the network quality performance value, vehicle speed performance value, and road congestion value are calculated;

[0007] Using the swarm optimization algorithm to obtain the adjustment coefficient combination k j , j = 1, 2, ..., 6; for the safety tasks, decision-making tasks, communication tasks and auxiliary tasks performed by the vehicle, the obtained network quality performance value, vehicle speed performance value and road congestion value are obtained, and the obtained adjustment coefficient combination k is applied at the same time j , respectively calculate the priority weight value of the safety task, the priority weight value of the decision task, the priority weight value of the communication task and the priority weight value of the auxiliary task;

[0008] Based on the obtained security task priority weight value, decision task priority weight value, communication task priority weight value and auxiliary task priority weight value, the computing resource allocation ratio and network resource allocation ratio of each task are further calculated; the computing resource allocation ratio and network resource allocation ratio of each task obtained are used to allocate computing resources and network resources respectively.

[0009] Preferably, the network quality performance value, the vehicle speed performance value and the road congestion degree value are calculated based on the acquired environmental data, and the specific operations are as follows:

[0010] Based on the bandwidth, delay and packet loss rate in the currently acquired network status data, a standard bandwidth range, a standard delay range and a standard packet loss rate range are set; based on the standard bandwidth range, the standard delay range and the standard packet loss rate range, the currently acquired bandwidth, delay and packet loss rate are linearly mapped to a range of 0 to 1, and a bandwidth normalization value a, a delay normalization value b and a packet loss rate normalization value c are obtained, and then a network quality performance value Q is calculated using the formula Q=0.3a+0.3(1-b)+0.4(1-c);

[0011] Based on the currently acquired vehicle speed data, a standard vehicle speed range is set, and based on the standard vehicle speed range, the currently acquired vehicle speed is linearly mapped to a range of 0 to 1 to obtain a vehicle speed performance value S;

[0012] Based on the currently acquired observation section vehicle data, the observation section vehicle data represents the number of vehicles in the fixed-length observation section where the vehicle is currently located, a standard number range is set, and the currently acquired number of vehicles is mapped to the range of 0 to 1 based on the standard number range to obtain the road congestion value T.

[0013] Preferably, the obtained network quality performance value, vehicle speed performance value and road congestion value are used, and the obtained adjustment coefficient combination k is applied simultaneously. j , respectively calculate the priority weight value of the safety task, the priority weight value of the decision task, the priority weight value of the communication task and the priority weight value of the auxiliary task. The specific operations are as follows:

[0014] Based on the obtained network quality performance value Q, vehicle speed performance value S and road congestion value T, for safety tasks, the formula V sec=1+k1S+k2T Calculate and obtain the security task priority weight value V sec , where k1 and k2 are the first adjustment coefficient and the second adjustment coefficient, respectively, and k1 and k2 respectively represent the influence of vehicle speed and road congestion on the safety task;

[0015] For decision-making tasks, use formula V dec =1+k3S+k4T Calculate and obtain the priority weight value V of the decision task dec , where k3 and k4 are the third adjustment coefficient and the fourth adjustment coefficient respectively, and k3 and k4 respectively represent the influence of vehicle speed and road congestion on the decision-making task;

[0016] For communication tasks, use formula V comm =1+k5Q+k6S Calculate and obtain the communication task priority weight value V comm , where k5 and k6 are the fifth adjustment coefficient and the sixth adjustment coefficient respectively, and k5 and k6 respectively represent the influence of network quality and vehicle speed on the communication task;

[0017] For auxiliary tasks, the auxiliary task priority weight value V aux Always 1.

[0018] Preferably, the specific operation of calculating the computing resource allocation ratio of each task is as follows:

[0019] For the allocation of computing resources, the initial computing allocation value of the security task is set to 4, the initial computing allocation value of the decision task is set to 3, the initial computing allocation value of the communication task is set to 2, and the initial computing allocation value of the auxiliary task is set to 1; then based on the obtained security task priority weight value V sec , decision-making task priority weight value V dec , communication task priority weight value V comm and auxiliary task priority weight value V aux , calculate and obtain the safety task calculation allocation value Decision task calculation allocation value Communication task calculation allocation value and auxiliary tasks to calculate the distribution value in, Using the formula Calculate the computing resource allocation ratio of security tasks, decision-making tasks, communication tasks and auxiliary tasks respectively i = 1, 2, 3, 4; and They represent the computing resource allocation ratios of safety tasks, decision-making tasks, communication tasks, and auxiliary tasks respectively.

[0020] Preferably, the specific operation of calculating the network resource allocation ratio of each task is as follows:

[0021] For the allocation of network resources, the initial network allocation value of the security task is set to 3, the initial network allocation value of the decision task is set to 2, the initial network allocation value of the communication task is set to 4, and the initial network allocation value of the auxiliary task is set to 1; based on the obtained security task priority weight value V sec , decision-making task priority weight value V dec , communication task priority weight value V comm and auxiliary task priority weight value V aux , calculate and obtain the safety task network allocation value Decision task network allocation value Communication Task Network Allocation Value and auxiliary task network assignment value in, Using the formula Calculate the network resource allocation ratios for security tasks, decision-making tasks, communication tasks, and auxiliary tasks respectively i = 1, 2, 3, 4; and Represents the network resource allocation ratio of safety task, decision task, communication task and auxiliary task respectively.

[0022] Preferably, obtain the adjustment coefficient combination k j The population optimization algorithm used is the genetic algorithm.

[0023] Preferably, the adjustment coefficient combination k is obtained by genetic algorithm j The specific operations are as follows:

[0024] Step 1: Set the population size, randomly generate x candidate coefficient combinations, each adjustment coefficient in the adjustment coefficient combination takes a value between 0 and 1, and form the randomly generated candidate coefficient combinations into an initial population, and set the maximum number of iterations;

[0025] Step 2: For any candidate coefficient combination, use the formula The fitness F of the candidate coefficient combination is calculated, where D represents the average processing delay of each task when the candidate coefficient combination is applied for resource allocation; R represents the resource utilization rate of the candidate coefficient combination for resource allocation, and ω1 and ω2 are both weight coefficients; this formula indicates that the lower the average processing delay of each task and the higher the resource utilization rate for resource allocation, the better the performance of the candidate coefficient combination;

[0026] Step 3: Select the candidate coefficient group with the highest current fitness as the elite individual of the current iteration number; then apply the roulette selection method to all candidate coefficient combinations except the elite individuals, select the corresponding number of candidate coefficient combinations with the best fitness performance according to the preset selection ratio, and form the selected candidate coefficient combinations into a selection individual set;

[0027] Step 4: Apply the preset crossover probability to perform a crossover operation on the candidate coefficient combinations in the selected individual set to generate a corresponding number of new candidate coefficient combinations, and form the obtained new candidate coefficient combinations into a crossover individual set; randomly select candidate coefficient combinations in the crossover individual set, and make small-range adjustments to the candidate coefficient combinations according to the preset mutation probability to obtain a mutation individual set;

[0028] Step 5: Merge the selected individual set and the mutated individual set to obtain the merged individual set, calculate the fitness of all candidate coefficient combinations in the merged individual set, and use the x-1 candidate coefficient combinations with the best fitness performance in the merged individual set and the selected elite individuals as the new generation population;

[0029] Step 6: Repeat steps 3 to 5 to iteratively update the population until the maximum number of iterations is reached, and use the candidate coefficient combination with the highest fitness in the last generation of the population as the adjustment coefficient combination k j .

[0030] Preferably, a computer-based network dynamic optimization system comprises:

[0031] The environmental assessment module includes a data monitoring unit and a calculation unit; the data monitoring unit is used to obtain the current environmental data of the vehicle at any monitoring time point during the driving process of the unmanned vehicle, and the environmental data includes network status data, vehicle speed data and observed road section vehicle data; the calculation unit is used to calculate the network quality performance value, vehicle speed performance value and road congestion value based on the acquired environmental data;

[0032] The task weight evaluation module includes an adjustment coefficient combination acquisition unit and a priority weight value calculation unit; the adjustment coefficient combination acquisition unit is used to obtain the adjustment coefficient combination k using a group optimization algorithm j The priority weight value calculation unit is used for the safety tasks, decision-making tasks, communication tasks and auxiliary tasks performed by the vehicle, and the obtained network quality performance value, vehicle speed performance value and road congestion value are obtained, and the obtained adjustment coefficient combination k is applied at the same time. j , respectively calculate the priority weight value of the safety task, the priority weight value of the decision task, the priority weight value of the communication task and the priority weight value of the auxiliary task;

[0033] The resource allocation module includes a resource allocation ratio calculation unit, a computing resource allocation unit and a network resource allocation unit; the resource allocation ratio calculation unit is used to further calculate the computing resource allocation ratio and the network resource allocation ratio of each task based on the acquired security task priority weight value, decision task priority weight value, communication task priority weight value and auxiliary task priority weight value; the computing resource allocation unit is used to allocate computing resources according to the acquired computing resource allocation ratio of each task; the network resource allocation unit is used to allocate network resources according to the acquired network resource allocation ratio of each task.

[0034] The present invention has the following advantages:

[0035] 1. The present invention can dynamically adjust the allocation of computing resources and network resources by real-time monitoring of the network status, vehicle speed and road congestion of the unmanned vehicle. The adaptive adjustment mechanism of this method enables the unmanned vehicle to instantly optimize the resource allocation strategy according to the real-time changing environmental data and task requirements; in the face of rapidly changing network conditions and traffic conditions, its excellent adaptability and real-time performance can ensure that the driving performance and decision-making efficiency of the unmanned vehicle are always maintained in the best state, thereby improving the vehicle's response speed and adaptability to complex traffic environments.

[0036] 2. The present invention obtains the combination of adjustment coefficients through genetic algorithms, and can then accurately calculate the priority weight value of each task, and allocate computing resources and network resources accordingly; reduce the average processing delay of tasks and improve resource utilization of resource allocation, thereby maximizing system efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 A schematic diagram of the structure of a network dynamic optimization system of a computer architecture adopted in an embodiment of the present invention. DETAILED DESCRIPTION

[0038] In order to enable persons skilled in the art to better understand the technical solutions in the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0039] Embodiment 1, a method for dynamic network optimization of a computer architecture, comprising:

[0040] At any monitoring time point during the driving process of the unmanned vehicle, the vehicle's current environmental data is obtained. The environmental data includes network status data, vehicle speed data and observed road section vehicle data; network status data includes key indicators such as bandwidth, delay and packet loss rate. These indicators reflect the current status of the vehicle's wireless communication network. The collection of network status data is crucial for evaluating and optimizing the communication between the vehicle and the control center, other vehicles and roadside infrastructure; network quality performance values, vehicle speed performance values ​​and road congestion values ​​are calculated based on the obtained environmental data; the purpose of this step is to convert the original environmental data into quantitative indicators that can be used for decision-making. These indicators can accurately reflect the current network and traffic conditions faced by the unmanned vehicle; through these quantitative performance values, the system can more accurately allocate resources and adjust task priorities, thereby optimizing the driving performance and decision-making efficiency of the unmanned vehicle;

[0041] Using the swarm optimization algorithm to obtain the adjustment coefficient combination k j , j = 1, 2, ..., 6; for the safety tasks, decision-making tasks, communication tasks and auxiliary tasks performed by the vehicle, safety tasks usually include emergency braking, obstacle avoidance, etc., which have extremely high requirements for real-time and reliability, so their priority is often set higher; decision-making tasks involve path planning, speed adjustment, etc., which require comprehensive consideration of multiple factors to ensure driving efficiency and safety; communication tasks include communication with the control center, other vehicles, etc., which are highly dependent on network quality; auxiliary tasks such as vehicle status monitoring and predictive maintenance, etc., although these tasks are important, they are not as urgent as other tasks; by obtaining the network quality performance value, vehicle speed performance value and road congestion value, and applying the obtained adjustment coefficient combination k j , respectively calculate the priority weight value of the safety task, the priority weight value of the decision task, the priority weight value of the communication task and the priority weight value of the auxiliary task; by calculating the priority weight values ​​of different tasks, the present invention can realize the dynamic allocation of computing resources and network resources; this dynamic allocation strategy ensures that key tasks (such as safety tasks and decision tasks) can be given priority when resources are tight, thereby improving the driving safety and decision-making efficiency of unmanned vehicles;

[0042] Based on the obtained safety task priority weight value, decision task priority weight value, communication task priority weight value and auxiliary task priority weight value, the computing resource allocation ratio and network resource allocation ratio of each task are further calculated; computing resources refer to the hardware and software resources used in the computer system to perform computing tasks, including but not limited to processors, memory, storage devices and graphics processing units; network resources refer to resources used for data transmission and communication in a network environment; the computing resource allocation ratio and network resource allocation ratio of each task obtained are used to allocate computing resources and network resources respectively; different allocation methods for computing resources and network resources are used to adapt to the specific needs and priorities of different tasks in unmanned vehicles for these resources, ensure that critical tasks such as safety and decision tasks can obtain sufficient computing power and network bandwidth to ensure real-time and reliability, and optimize the resource usage of non-critical tasks such as auxiliary tasks to improve the efficiency and performance of the overall system; for example, communication tasks may require continuous network connection and high bandwidth to ensure real-time transmission of data, while computing-intensive tasks such as environmental perception and path planning may require more computing resources to process large amounts of sensor data. Through differentiated resource allocation strategies, driverless vehicles can respond more flexibly and efficiently to dynamically changing driving environments and mission requirements, thereby improving driving safety and decision-making accuracy.

[0043] The network quality performance value, vehicle speed performance value and road congestion value are calculated based on the acquired environmental data. The specific operations are as follows:

[0044] Based on the bandwidth, delay and packet loss rate in the currently acquired network status data, a standard bandwidth range, a standard delay range and a standard packet loss rate range are set; based on the standard bandwidth range, the standard delay range and the standard packet loss rate range, the currently acquired bandwidth, delay and packet loss rate are linearly mapped to a range of 0 to 1, and a bandwidth normalization value a, a delay normalization value b and a packet loss rate normalization value c are obtained, and then a network quality performance value Q is calculated using the formula Q=0.3a+0.3(1-b)+0.4(1-c);

[0045] Based on the currently acquired vehicle speed data, a standard vehicle speed range is set, and based on the standard vehicle speed range, the currently acquired vehicle speed is linearly mapped to a range of 0 to 1 to obtain a vehicle speed performance value S;

[0046] Based on the currently acquired observation section vehicle data, the observation section vehicle data represents the number of vehicles in the fixed-length observation section where the vehicle is currently located, a standard number range is set, and the currently acquired number of vehicles is mapped to the range of 0 to 1 based on the standard number range to obtain the road congestion value T.

[0047] By obtaining the network quality performance value, vehicle speed performance value and road congestion value, the obtained adjustment coefficient combination k is applied at the same timej , respectively calculate the priority weight value of the safety task, the priority weight value of the decision task, the priority weight value of the communication task and the priority weight value of the auxiliary task. The specific operations are as follows:

[0048] Based on the obtained network quality performance value Q, vehicle speed performance value S and road congestion value T, for safety tasks, the formula V sec =1+k1S+k2T Calculate and obtain the security task priority weight value V sec , where k1 and k2 are the first adjustment coefficient and the second adjustment coefficient, respectively, and k1 and k2 respectively represent the influence of vehicle speed and road congestion on the safety task;

[0049] For decision-making tasks, use formula V dec =1+k3S+k4T Calculate and obtain the priority weight value V of the decision task dec , where k3 and k4 are the third adjustment coefficient and the fourth adjustment coefficient respectively, and k3 and k4 respectively represent the influence of vehicle speed and road congestion on the decision-making task;

[0050] For communication tasks, use formula V comm =1+k5Q+k6S Calculate and obtain the communication task priority weight value V comm , where k5 and k6 are the fifth adjustment coefficient and the sixth adjustment coefficient respectively, and k5 and k6 respectively represent the influence of network quality and vehicle speed on the communication task;

[0051] For auxiliary tasks, the auxiliary task priority weight value V aux Always 1.

[0052] The specific operations for calculating the computing resource allocation ratio of each task are as follows:

[0053] For the allocation of computing resources, the initial computing allocation value of the security task is set to 4, the initial computing allocation value of the decision task is set to 3, the initial computing allocation value of the communication task is set to 2, and the initial computing allocation value of the auxiliary task is set to 1; then based on the obtained security task priority weight value V sec , decision-making task priority weight value V dec , communication task priority weight value V comm and auxiliary task priority weight value V aux , calculate and obtain the safety task calculation allocation value Decision task calculation allocation value Communication task calculation allocation value and auxiliary tasks to calculate the distribution value in, Using the formula Calculate the computing resource allocation ratio of security tasks, decision-making tasks, communication tasks and auxiliary tasks respectively i = 1, 2, 3, 4; and They represent the computing resource allocation ratios of safety tasks, decision-making tasks, communication tasks, and auxiliary tasks respectively.

[0054] The specific operations for calculating the network resource allocation ratio of each task are as follows:

[0055] For the allocation of network resources, the initial network allocation value of the security task is set to 3, the initial network allocation value of the decision task is set to 2, the initial network allocation value of the communication task is set to 4, and the initial network allocation value of the auxiliary task is set to 1; based on the obtained security task priority weight value V sec , decision-making task priority weight value V dec , communication task priority weight value V comm and auxiliary task priority weight value V aux , calculate and obtain the safety task network allocation value Decision task network allocation value Communication Task Network Allocation Value and auxiliary task network assignment value in, Using the formula Calculate the network resource allocation ratios for security tasks, decision-making tasks, communication tasks, and auxiliary tasks respectively i = 1, 2, 3, 4; and Represents the network resource allocation ratio of safety task, decision task, communication task and auxiliary task respectively.

[0056] Get adjustment coefficient combination k j The population optimization algorithm used is the genetic algorithm.

[0057] Obtaining adjustment coefficient combination k through genetic algorithm j The specific operations are as follows:

[0058] Step 1: Set the population size, randomly generate x candidate coefficient combinations, each adjustment coefficient in the adjustment coefficient combination takes a value between 0 and 1, and form the randomly generated candidate coefficient combinations into an initial population, and set the maximum number of iterations;

[0059] Step 2: For any candidate coefficient combination, use the formula The fitness F of the candidate coefficient combination is calculated, where D represents the average processing delay of each task when the candidate coefficient combination is applied for resource allocation; R represents the resource utilization rate of the candidate coefficient combination for resource allocation, and ω1 and ω2 are both weight coefficients; this formula indicates that the lower the average processing delay of each task and the higher the resource utilization rate for resource allocation, the better the performance of the candidate coefficient combination;

[0060] Step 3: Select the candidate coefficient group with the highest current fitness as the elite individual of the current iteration number; then apply the roulette selection method to all candidate coefficient combinations except the elite individuals, select the corresponding number of candidate coefficient combinations with the best fitness performance according to the preset selection ratio, and form the selected candidate coefficient combinations into a selection individual set;

[0061] Step 4: Apply the preset crossover probability to perform a crossover operation on the candidate coefficient combinations in the selected individual set to generate a corresponding number of new candidate coefficient combinations, and form the obtained new candidate coefficient combinations into a crossover individual set; randomly select candidate coefficient combinations in the crossover individual set, and make small-range adjustments to the candidate coefficient combinations according to the preset mutation probability to obtain a mutation individual set;

[0062] Step 5: Merge the selected individual set and the mutated individual set to obtain the merged individual set, calculate the fitness of all candidate coefficient combinations in the merged individual set, and use the x-1 candidate coefficient combinations with the best fitness performance in the merged individual set and the selected elite individuals as the new generation population;

[0063] Step 6: Repeat steps 3 to 5 to iteratively update the population until the maximum number of iterations is reached, and use the candidate coefficient combination with the highest fitness in the last generation of the population as the adjustment coefficient combination k j .

[0064] Embodiment 2, a computer-based network dynamic optimization system, such as Figure 1 As shown, including:

[0065] The environmental assessment module includes a data monitoring unit and a calculation unit; the data monitoring unit is used to obtain the current environmental data of the vehicle at any monitoring time point during the driving process of the unmanned vehicle, and the environmental data includes network status data, vehicle speed data and observed road section vehicle data; the calculation unit is used to calculate the network quality performance value, vehicle speed performance value and road congestion value based on the acquired environmental data;

[0066] The task weight evaluation module includes an adjustment coefficient combination acquisition unit and a priority weight value calculation unit; the adjustment coefficient combination acquisition unit is used to obtain the adjustment coefficient combination k using a group optimization algorithm jThe priority weight value calculation unit is used for the safety tasks, decision-making tasks, communication tasks and auxiliary tasks performed by the vehicle, and the obtained network quality performance value, vehicle speed performance value and road congestion value are obtained, and the obtained adjustment coefficient combination k is applied at the same time. j , respectively calculate the priority weight value of the safety task, the priority weight value of the decision task, the priority weight value of the communication task and the priority weight value of the auxiliary task;

[0067] The resource allocation module includes a resource allocation ratio calculation unit, a computing resource allocation unit and a network resource allocation unit; the resource allocation ratio calculation unit is used to further calculate the computing resource allocation ratio and the network resource allocation ratio of each task based on the acquired security task priority weight value, decision task priority weight value, communication task priority weight value and auxiliary task priority weight value; the computing resource allocation unit is used to allocate computing resources according to the acquired computing resource allocation ratio of each task; the network resource allocation unit is used to allocate network resources according to the acquired network resource allocation ratio of each task.

[0068] It should be understood that those skilled in the art can make improvements or changes based on the above description, and all such improvements and changes should fall within the scope of protection of the appended claims of the present invention. Parts not described in detail in this specification belong to the prior art known to those skilled in the art.

Claims

1. A method for dynamic network optimization of computer architecture, characterized in that: include: At any monitoring time point during the driving process of the unmanned vehicle, the vehicle's current environmental data is obtained, including network status data, vehicle speed data, and observed road section vehicle data; based on the obtained environmental data, the network quality performance value, vehicle speed performance value, and road congestion value are calculated; Using the swarm optimization algorithm to obtain the adjustment coefficient combination k j , j = 1, 2, ..., 6; For the safety tasks, decision-making tasks, communication tasks and auxiliary tasks performed by the vehicle, the obtained network quality performance value, vehicle speed performance value and road congestion value are used, and the obtained adjustment coefficient combination k is applied at the same time. j , respectively calculate the priority weight value of the safety task, the priority weight value of the decision task, the priority weight value of the communication task and the priority weight value of the auxiliary task; Based on the obtained security task priority weight value, decision task priority weight value, communication task priority weight value and auxiliary task priority weight value, the computing resource allocation ratio and network resource allocation ratio of each task are further calculated; the computing resource allocation ratio and network resource allocation ratio of each task obtained are used to allocate computing resources and network resources respectively.

2. A method for dynamic network optimization of computer architecture according to claim 1, characterized in that: The network quality performance value, vehicle speed performance value and road congestion value are calculated based on the acquired environmental data. The specific operations are as follows: Based on the bandwidth, delay and packet loss rate in the currently acquired network status data, a standard bandwidth range, a standard delay range and a standard packet loss rate range are set; based on the standard bandwidth range, the standard delay range and the standard packet loss rate range, the currently acquired bandwidth, delay and packet loss rate are linearly mapped to a range of 0 to 1, and a bandwidth normalization value a, a delay normalization value b and a packet loss rate normalization value c are obtained, and then a network quality performance value Q is calculated using the formula Q=0.3a+0.3(1-b)+0.4(1-c); Based on the currently acquired vehicle speed data, a standard vehicle speed range is set, and based on the standard vehicle speed range, the currently acquired vehicle speed is linearly mapped to a range of 0 to 1 to obtain a vehicle speed performance value S; Based on the currently acquired observation section vehicle data, the observation section vehicle data represents the number of vehicles in the fixed-length observation section where the vehicle is currently located, a standard number range is set, and the currently acquired number of vehicles is mapped to the range of 0 to 1 based on the standard number range to obtain the road congestion value T.

3. A method for dynamic network optimization of computer architecture according to claim 2, characterized in that: By obtaining the network quality performance value, vehicle speed performance value and road congestion value, the obtained adjustment coefficient combination k is applied at the same time j , respectively calculate the priority weight value of the safety task, the priority weight value of the decision task, the priority weight value of the communication task and the priority weight value of the auxiliary task. The specific operations are as follows: Based on the obtained network quality performance value Q, vehicle speed performance value S and road congestion value T, for safety tasks, the formula V sec =1+k1S+k2T Calculate and obtain the security task priority weight value V sec , where k1 and k2 are the first adjustment coefficient and the second adjustment coefficient, respectively, and k1 and k2 respectively represent the influence of vehicle speed and road congestion on the safety task; For decision-making tasks, use formula V dec =1+k3S+k4T Calculate and obtain the priority weight value V of the decision task dec , where k3 and k4 are the third adjustment coefficient and the fourth adjustment coefficient respectively, and k3 and k4 respectively represent the influence of vehicle speed and road congestion on the decision-making task; For communication tasks, use formula V comm =1+k5Q+k6S Calculate and obtain the communication task priority weight value V comm , where k5 and k6 are the fifth adjustment coefficient and the sixth adjustment coefficient respectively, and k5 and k6 respectively represent the influence of network quality and vehicle speed on the communication task; For auxiliary tasks, the auxiliary task priority weight value V aux Always 1.

4. A method for dynamic network optimization of computer architecture according to claim 3, characterized in that: The specific operations for calculating the computing resource allocation ratio of each task are as follows: For the allocation of computing resources, the initial computing allocation value of the security task is set to 4, the initial computing allocation value of the decision task is set to 3, the initial computing allocation value of the communication task is set to 2, and the initial computing allocation value of the auxiliary task is set to 1; then based on the obtained security task priority weight value V sec , decision-making task priority weight value V dec , communication task priority weight value V comm and auxiliary task priority weight value V aux , calculate and obtain the safety task calculation allocation value Decision task calculation allocation value Communication task calculation allocation value and auxiliary tasks to calculate the distribution value in, Using the formula Calculate the computing resource allocation ratio of security tasks, decision-making tasks, communication tasks and auxiliary tasks respectively and They represent the computing resource allocation ratios of safety tasks, decision-making tasks, communication tasks, and auxiliary tasks respectively.

5. A method for dynamic network optimization of computer architecture according to claim 4, characterized in that: The specific operations for calculating the network resource allocation ratio of each task are as follows: For the allocation of network resources, the initial network allocation value of the security task is set to 3, the initial network allocation value of the decision task is set to 2, the initial network allocation value of the communication task is set to 4, and the initial network allocation value of the auxiliary task is set to 1; based on the obtained security task priority weight value V sec , decision-making task priority weight value V dec , communication task priority weight value V comm and auxiliary task priority weight value V aux , calculate and obtain the safety task network allocation value Decision task network allocation value Communication Task Network Allocation Value and auxiliary task network assignment value in, Using the formula Calculate the network resource allocation ratios for security tasks, decision-making tasks, communication tasks, and auxiliary tasks respectively and Represents the network resource allocation ratio of safety task, decision task, communication task and auxiliary task respectively.

6. A method for dynamic network optimization of computer architecture according to claim 5, characterized in that: Get adjustment coefficient combination k j The population optimization algorithm used is the genetic algorithm.

7. A method for dynamic network optimization of computer architecture according to claim 6, characterized in that: Obtaining adjustment coefficient combination k through genetic algorithm j The specific operations are as follows: Step 1: Set the population size, randomly generate x candidate coefficient combinations, each adjustment coefficient in the adjustment coefficient combination takes a value between 0 and 1, and form the randomly generated candidate coefficient combinations into an initial population, and set the maximum number of iterations; Step 2: For any candidate coefficient combination, use the formula The fitness F of the candidate coefficient combination is calculated, where D represents the average processing delay of each task when the candidate coefficient combination is applied for resource allocation; R represents the resource utilization rate of the candidate coefficient combination for resource allocation, and ω1 and ω2 are both weight coefficients; this formula indicates that the lower the average processing delay of each task and the higher the resource utilization rate for resource allocation, the better the performance of the candidate coefficient combination; Step 3: Select the candidate coefficient group with the highest current fitness as the elite individual of the current iteration number; then apply the roulette selection method to all candidate coefficient combinations except the elite individuals, select the corresponding number of candidate coefficient combinations with the best fitness performance according to the preset selection ratio, and form the selected candidate coefficient combinations into a selection individual set; Step 4: Apply the preset crossover probability to perform a crossover operation on the candidate coefficient combinations in the selected individual set to generate a corresponding number of new candidate coefficient combinations, and form the obtained new candidate coefficient combinations into a crossover individual set; randomly select candidate coefficient combinations in the crossover individual set, and make small-range adjustments to the candidate coefficient combinations according to the preset mutation probability to obtain a mutation individual set; Step 5: Merge the selected individual set and the mutated individual set to obtain the merged individual set, calculate the fitness of all candidate coefficient combinations in the merged individual set, and use the x-1 candidate coefficient combinations with the best fitness performance in the merged individual set and the selected elite individuals as the new generation population; Step 6: Repeat steps 3 to 5 to iteratively update the population until the maximum number of iterations is reached, and use the candidate coefficient combination with the highest fitness in the last generation of the population as the adjustment coefficient combination k j .

8. A computer-based network dynamic optimization system, characterized in that: The system is applied to a method for dynamic network optimization of a computer architecture as described in any one of claims 1 to 7, comprising: The environmental assessment module includes a data monitoring unit and a calculation unit; the data monitoring unit is used to obtain the current environmental data of the vehicle at any monitoring time point during the driving process of the unmanned vehicle, and the environmental data includes network status data, vehicle speed data and observed road section vehicle data; the calculation unit is used to calculate the network quality performance value, vehicle speed performance value and road congestion value based on the acquired environmental data; The task weight evaluation module includes an adjustment coefficient combination acquisition unit and a priority weight value calculation unit; the adjustment coefficient combination acquisition unit is used to obtain the adjustment coefficient combination k using a group optimization algorithm j The priority weight value calculation unit is used for the safety tasks, decision-making tasks, communication tasks and auxiliary tasks performed by the vehicle, and the obtained network quality performance value, vehicle speed performance value and road congestion value are obtained, and the obtained adjustment coefficient combination k is applied at the same time. j , respectively calculate the priority weight value of the safety task, the priority weight value of the decision task, the priority weight value of the communication task and the priority weight value of the auxiliary task; The resource allocation module includes a resource allocation ratio calculation unit, a computing resource allocation unit and a network resource allocation unit; the resource allocation ratio calculation unit is used to further calculate the computing resource allocation ratio and the network resource allocation ratio of each task based on the acquired security task priority weight value, decision task priority weight value, communication task priority weight value and auxiliary task priority weight value; the computing resource allocation unit is used to allocate computing resources according to the acquired computing resource allocation ratio of each task; the network resource allocation unit is used to allocate network resources according to the acquired network resource allocation ratio of each task.

Images