A System Intelligent Decision-Making Method for a Multi-Domain Integrated Platform

By employing a system intelligent decision-making method based on a multi-domain integrated platform, the problem of parameter analysis lag in unmanned platforms under extreme environments was solved. This enabled rapid and accurate calculation of state characteristics, improved the system's self-learning and adaptability, and ensured operational stability and reliability.

CN119046869BActive Publication Date: 2026-11-17CHINA NORTH VEHICLE RES INST
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411069258.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-06
Publication Date
2026-11-17
Estimated Expiration
2044-08-06

AI Technical Summary

Technical Problem

Unmanned platforms struggle to perform rapid and accurate parameter analysis and integrated parameter calculations consistent with state characteristics in extreme environments, resulting in low system computational efficiency and delayed changes in the dynamic environment, which affects system performance stability.

Method used

The system intelligent decision-making method adopts a multi-domain integrated platform. Through system state preprocessing, multi-domain system state coupling calculation, and state-coupled calculation decision, including data state prefabrication, multi-system data state preprocessing, parameter coupling calculation of power domain, chassis domain and intelligent driving domain, as well as parameter hierarchical processing and system decision classification, it realizes the selection and coupling calculation of characteristic state variables of different subdomains.

Benefits of technology

It enhances the self-learning and intelligent adaptation capabilities of the UAV platform under various operating conditions, improves the efficiency and accuracy of system decision-making, and ensures the stability and reliability of operation in uncertain environments.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119046869B_ABST
    Figure CN119046869B_ABST
Patent Text Reader

Abstract

This invention belongs to the field of unmanned vehicle swarm control technology, specifically involving a system intelligent decision-making method for a multi-domain integrated platform. It comprises three parts: system state preprocessing, multi-domain system state coupling calculation, and computational decision-making under state coupling. These three parts form the framework and architecture of the system intelligent decision-making. The foundation of intelligent decision-making includes vertical multi-dimensional data feature extraction and vertical multi-domain collaboration and parameter coupling. Data feature extraction involves different sub-domain state pre-processing and multi-parameter state coupling calculation processes, corresponding to the system state preprocessing and multi-domain system state coupling calculation processes. Multi-domain collaboration completes hierarchical processing and decision classification based on multi-domain coupling parameters, synchronously corresponding to the computational decision-making module under state coupling. Through multi-directional and multi-dimensional data iteration, the system state transmission and judgment and control command decision-making are completed, achieving integrated processing of system data under integrated geographical and operational conditions, and completing the online intelligent decision-making process of the multi-domain platform in the form of a data cluster.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of unmanned vehicle cluster control technology, specifically relating to a system intelligent decision-making method for a multi-domain integrated platform. Background Technology

[0002] In extreme environments such as swamps, deserts, frigid zones, plateaus, and oceans, unmanned platforms need to adapt to these environments and replace human labor to complete actual work requirements. In such cases, detailed and accurate environmental parameters may be inaccurate. Therefore, during real-time operation, unmanned mobile platforms need to learn and make judgments based on data, gradually acquiring the ability for system self-adaptation and parameter self-acquisition and self-analysis, thereby fulfilling the mobility and functional requirements of the target scenario.

[0003] The unmanned platform itself is modularly designed, composed of different systems, and its internal information transmission and processing are quite complex. To ensure strong environmental adaptability, the unmanned platform needs sufficient environmental information input and powerful computational analysis capabilities, based on the effective coordination of internal information. This allows it to fully utilize existing data for analysis and perform effective information calculation and feature extraction based on changes in its own state information, thereby completing the identification and positioning of the current environment and the platform itself.

[0004] However, the input of external environmental information is inseparable from the high-quality requirements of sensors on unmanned intelligent platforms. But due to limitations such as vehicle space, weight, and manufacturing costs, the number and arrangement of sensors are also subject to certain constraints, thus limiting their ability to perceive external information. Under these circumstances, considering both objective constraints and performance requirements, analyzing the dynamic changes in vehicle status in reverse to understand the patterns of environmental factor changes and system state changes becomes a realistic and feasible solution.

[0005] The computing power of chips installed in actual vehicles has its limits. Based on the existing vehicle data stored within the vehicle, it is not possible to select and analyze all states without restriction. This is because different states may have inter-state coupling relationships. While this approach is simple, the actual operation would place a significant burden on the system, thereby reducing its computational efficiency. On the other hand, processing such massive amounts of data would also prevent the acquisition of effective target information in a short time. Ineffective or delayed computation corresponds to the system's lagging response to dynamic environmental changes, leading to system performance instability and uncertainty.

[0006] Therefore, selecting effective states based on existing state data, and quickly and accurately performing parameter analysis and integrated parameter calculation consistent with state characteristics across multiple states, has become a key challenge in the multi-domain collaborative decision-making process of unmanned platforms. Furthermore, identifying coupling parameters, analyzing state variables, and making intelligent manipulations and rational decisions based on the analysis results are also crucial technical challenges that need to be addressed during multi-domain collaborative operation. Summary of the Invention

[0007] (a) Technical problems to be solved

[0008] The technical problem to be solved by this invention is: how to provide a system intelligent decision-making method for a multi-domain integrated platform.

[0009] (II) Technical Solution

[0010] To address the aforementioned technical problems, this invention provides a system intelligent decision-making method for a multi-domain integrated platform, the method comprising:

[0011] Step 1: System state preprocessing;

[0012] Step 2: Calculation of state coupling in multi-domain systems;

[0013] Step 3: Computational decision-making under state coupling.

[0014] Step 1 includes:

[0015] Step 11: Data status pre-configuration;

[0016] Step 12: Multi-system data status preprocessing.

[0017] Step 2 includes:

[0018] Step 21: Coupled calculation of dynamic domain parameters;

[0019] Step 22: Chassis domain parameter coupling calculation;

[0020] Step 23: Coupled calculation of intelligent driving domain parameters.

[0021] Step 3 includes:

[0022] Step 31: Parameter hierarchical processing;

[0023] Step 32: System decision classification;

[0024] Step 33: Control command decision.

[0025] (III) Beneficial Effects

[0026] The intelligent decision-making method of this invention mainly involves data feature extraction, hierarchical classification and recognition, and online comprehensive evaluation under system state input conditions. During actual vehicle operation, disturbance information, sensor data, and multi-domain information interact, requiring reasonable state selection based on system performance requirements. Secondly, it necessitates in-depth mining and online analysis of existing input data, completing a quantitative macro-evaluation of data from different domains and multiple data points under hierarchical quantification. The quantitative evaluation results serve as an intermediate transition layer input to the upper decision layer, providing effective data references for upper-level system decision-making under specific, unknown, and complex operating conditions. This enables the self-learning and intelligent adaptation capabilities of the unmanned aerial vehicle platform, thereby enhancing its practical application and expansion capabilities across multiple scenarios.

[0027] In view of the problems of existing technologies, this invention proposes a system intelligent decision-making method for a multi-domain integrated platform based on the selection of target state variables according to the functional differences of multiple system domains. This method selects characteristic state variables from different subdomains, and completes the transition from single information to subdomain information through variable normalization and coupled calculation. Furthermore, by performing gridded analysis and judgment on the coupling parameters of different subdomains, it completes parameter hierarchical processing and system decision classification, thereby realizing subdomain state analysis and judgment based on quantitative intervals, and simultaneously completing the system's primary decision-making and real-time intervention control, providing a practical and feasible solution for the stability and reliability of system operation under uncertain environments.

[0028] Compared with existing methods, this invention considers vertical calculation from underlying data to core indicator data and completes vertical segmentation of multi-domain information. Based on a three-dimensional data architecture, it performs multi-directional analysis of the system, thereby completing bidirectional cyclic layering of forward networked analysis and reverse quantitative control transmission. This completes the connection between data and the overall system at the system level, realizing the organic integration of online analysis and command transmission processes, which has significant implications for improving decision-making efficiency and accuracy. Attached Figure Description

[0029] Figure 1 This is a diagram of the multi-domain integrated system architecture for unmanned platforms.

[0030] Figure 2 This is a flowchart of the intelligent decision-making process for a multi-domain integrated platform. Detailed Implementation

[0031] To make the objectives, contents, and advantages of the present invention clearer, the specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples.

[0032] To address the problems of existing technologies, this invention provides a system intelligent decision-making method for a multi-domain integrated platform, the method comprising:

[0033] Step 1: System state preprocessing;

[0034] Step 2: Calculation of state coupling in multi-domain systems;

[0035] Step 3: Computational decision-making under state coupling.

[0036] Specifically, step 1, system state preprocessing, includes:

[0037] Step 11: Data status pre-configuration;

[0038] Step 12: Multi-system data status preprocessing.

[0039] The data state preset in step 11 is as follows:

[0040] Within different subdomains of the platform, target state variables are selected, and the selected system variables must meet the requirements of characterizing the performance of the vehicle's modular subsystems.

[0041] In the power domain, parameters related to the characteristics of the vehicle's power system are selected for calculation and analysis; in the chassis domain, characteristic parameters related to the suspension state, drive / braking state, and steering subsystem dynamics of the vehicle chassis are selected for calculation and analysis; and in the intelligent driving domain, vehicle and external parameters related to multi-vehicle intelligent collaboration are selected for comprehensive judgment.

[0042] The specific selection results are as follows: In the power domain, the vehicle speed v, the vehicle power demand P, and the real-time battery SOC are selected as state variables; in the chassis domain, the vehicle vertical displacement d is selected. v Braking force F b Driving force F d Steering angle δ is used as a state variable; within the intelligent driving domain, the maximum longitudinal speed duration interval i is selected. vl Horizontal and longitudinal maximum vehicle speed duration range i vc And the longitudinal distance s from the nearest vehicle l and lateral distance s c As a state variable; where the vehicle speed duration interval i vl and i vc This refers to the longest continuous driving time of a vehicle at its current maximum speed. Together with the distance between the vehicle and its two adjacent vehicles, it represents the safety between the two closest vehicles during multi-vehicle travel.

[0043] The multi-system data status preprocessing in step 12 is as follows:

[0044] Based on the selected state variables, the specific intervals corresponding to the selected state variables in different domains are first defined, and the state variables in different subdomains are then normalized and quantized.

[0045] Within the power domain, the corresponding ranges for vehicle speed v, vehicle power demand P, and current battery SOC are: [v min ,v max ]、[P min ,P max [SOC] min SOC max At this point, the relevant parameters within the dynamic domain are normalized, and the normalized state parameters corresponding to different parameters are as follows:

[0046]

[0047] Vertical displacement d of the vehicle body within the chassis domain v Braking force F b Driving force F d The interval ranges corresponding to the steering angle δ are respectively: [d v_min ,d v_max ]、[F b_min ,F b_max ]、[F d_min ,F d_max ]、[δ min ,δ max At this point, the relevant parameters within the chassis domain are normalized. The normalized state parameters corresponding to different parameters are as follows:

[0048]

[0049]

[0050] Maximum longitudinal speed duration i in intelligent driving domain vl Maximum lateral and longitudinal speed duration i vc Longitudinal vehicle spacing s l and lateral distance s c The corresponding safe interval ranges are: [i vl_min i vl_max ]、[i vc_min i vc_max ]、[s l_min ,s l_max ]、[s c_min ,s c_max At this point, the relevant parameters within the intelligent driving domain are normalized. The normalized state parameters corresponding to different parameters are as follows:

[0051]

[0052]

[0053] Among them, i vl_max and i vc_maxThese refer to the longest time a vehicle travels at its maximum longitudinal and lateral speeds, respectively. vl_min and i vc_min These refer to the shortest time a vehicle can travel at its minimum longitudinal and lateral speeds, respectively, and are generally taken as 0; however, if the lateral or longitudinal distance between the nearest vehicle and the current vehicle is greater than the safe interval [s], then... l_min ,s l_max ] and [s c_min ,s c_max ], then at this time the parameter N sl and N sc Then the corresponding value is 1.

[0054] In step 2, multi-domain system state coupling calculation is performed based on the normalized state parameters in different platform subdomains obtained in step 1.

[0055] Step 2 includes:

[0056] Step 21: Coupled calculation of dynamic domain parameters;

[0057] Step 22: Chassis domain parameter coupling calculation;

[0058] Step 23: Coupled calculation of intelligent driving domain parameters.

[0059] Specifically, in step 21, the dynamic domain parameter coupling calculation is as follows:

[0060] For N in the dynamic domain v N P and N SOC These three normalized state parameters are used to perform parameter coupling calculations within the dynamic domain;

[0061] Within the power domain, the vehicle's power P is primarily determined by the actual driving speed; therefore, vehicle speed v and power P are closely related, and redundant power calculations must be avoided during the calculation process. The actual coupled calculation process requires consideration of N... v and N P The size of N is determined and calculated; at the same time, based on the range of battery SOC, N is also... SOC The value is divided into the limiting working interval [0, α]. l ) and adaptable work range [α] l [1], where the working interval boundary value α l Take values ​​within the range of [20%, 25%];

[0062] Define the exponential critical value α1, where α1 takes values ​​in the range [0.45, 0.55], and its corresponding values ​​are as follows:

[0063]

[0064] Further define the power boundary interval value α1, if N v ≥α1, and N P If ≥α1, then the system coupling parameter N1 in the dynamic domain is:

[0065]

[0066] If N v ≥α1, and N P <α1; or N v <α1, and N P If ≥α1, then the system coupling parameter N1 in the dynamic domain is:

[0067]

[0068] If N v <α1, and N P If α1 < α1, then the system coupling parameter N1 in the dynamic domain is:

[0069]

[0070] Specifically, in step 22, the chassis domain parameter coupling calculation is as follows:

[0071] For N in the chassis domain dv N Fb N Fd and N δ These four normalized state parameters are used to perform parameter coupling calculations within the chassis domain;

[0072] During the calculation, since the driving force and braking force are related, it is necessary to first perform state coupling calculations; the coupling betweenness δ is introduced, and the specific coupling parameter N is... db The calculation equation is as follows:

[0073]

[0074] In the formula, the coupling intermediate value δ is a nonlinear function, and its calculation formula is as follows:

[0075]

[0076] Based on this, further research on N dv N db and N δ These three parameters are coupled for calculation; considering that different parameters correspond to the actual state of the vehicle chassis system, the magnitude of the three numbers is determined as follows:

[0077]

[0078] In the formula, the function median(N)dv N db N δ () represents the middle value among these three values;

[0079] The system coupling parameter N2 within the chassis domain is calculated as follows:

[0080]

[0081] In step 23, the intelligent driving domain parameter coupling calculation is as follows:

[0082] Based on the normalized state parameter N in the intelligent driving domain ivl N ivc N sl and N sc Perform system state coupling calculations within the intelligent driving domain;

[0083] Within the intelligent driving domain, the longitudinal and lateral distances and actual speeds between adjacent vehicles determine the dynamic safety of multi-vehicle driving. Therefore, the normalized parameters are first determined according to N. ivl N sl and N ivc N sc The calculations are performed in pairs; first, the longitudinal and lateral safety factors N between the two vehicles are calculated separately. pl and N pc Then the calculation formula is as follows:

[0084]

[0085] Next, the parameters are determined, and process parameter N is defined. 3p1 and N 3p2 :

[0086] (1) If N pi =0, k=l,c, then use N 3p1 =0 and N 3p2 If the result is 0, proceed to the next step of the calculation.

[0087] (2) If N pk =1, k=l,c, at this time N ivk =N sk Then using N 3p1 =N ivl =N sl and N 3p2 =N ivc =N sc Proceed to the next step of the determination calculation;

[0088] (3) If N pk >1, k=l,c, at this time Nivk <N sk Then using N 3p1 =N sl and N 3p2 =N sc Proceed to the next step of the determination calculation;

[0089] (4) If N pk <1, k = l, c, at this time N ivk >N sk Then we need to calculate N. 3p1 and N 3p2 The corresponding calculation equation at this time is:

[0090]

[0091] If the calculated N 3pj ≤1, j=1,2, then N 3pj Take the actual calculated value; if N 3pj >1, then N 3pj Take 1;

[0092] N obtained using the above calculation process 3p1 and N 3p2 Further calculation of the system coupling parameter N3 within the intelligent driving domain; considering the dynamic distance changes of the vehicle during actual driving, the actual coupling parameter N3 can be obtained as follows:

[0093]

[0094] In step 3, the computational decision-making under state coupling includes:

[0095] Step 31: Parameter hierarchical processing;

[0096] Step 32: System decision classification;

[0097] Step 33: Control command decision.

[0098] In step 31, the system parameters are graded based on the calculated coupling parameters N1, N2, and N3.

[0099] Since the coupling parameters N1, N2, and N3 fall within the interval [0,1], they are divided into four intervals: [0,β1], [β1,β2], [β2,β3], and [β3,1], representing the four processes in the actual decision-making process: system initiation, normal response, control intervention, and risk warning. The values ​​of coupling parameters N1, N2, and N3 are related to priority; larger values ​​indicate higher urgency. The interval boundaries have the following relationship: 0 < β1 < β2 < β3 < 1.

[0100] Based on the interval settings, the value of the interval threshold β0 needs to be determined first; the calculation of β0 mainly considers the difference between the maximum and minimum values ​​of the coupling parameters in different subdomains of the current system. Approaching the critical upper limit value The degree of difference between the maximum value and the critical value; based on its specific magnitude, the dynamic value of β0 can be determined as follows:

[0101]

[0102] Therefore, based on the value of the interval threshold β0, the corresponding values ​​of β1, β2, and β3 can be calculated as follows:

[0103]

[0104] In step 32, based on the calculated coupling parameters N1, N2, and N3, and the results of parameter hierarchical processing, the system decision classification is further carried out.

[0105] Based on the calculated coupling parameters N1, N2, N3, and interval boundary values ​​β1, β2, and β3, a multi-level classification system is determined. Based on the multi-domain coupling data, it is determined whether the corresponding region needs intervention, thereby obtaining the control domain priority L of the system decision. The specific determination process is shown in Table 1.

[0106] Table 1 System Multilevel Classification Process

[0107]

[0108]

[0109] The system decision priority values ​​in the table are related to the degree of priority. A value of 0 indicates that the system is normal and no action is required. The larger the value, the higher the urgency of the action.

[0110] Meanwhile, a flag F is set to locate the parameter quantity within the danger warning area, with an initial value of 0; when any of the coupling parameters N1, N2 and N3 is in the [β3,1] region, the value of F is incremented by 1, and the results are shown in Table 2.

[0111] Table 2 Statistics of System Warning Areas

[0112]

[0113] In step 33, the system control command decision-making process is carried out based on the boundary values ​​β1, β2, β3, system decision priority L, and the number of danger warning areas F.

[0114] If L=0 and F=0, the system will operate normally passively and will not require intervention or control via commands.

[0115] If 1≤L≤3, F=0. At this time, the system control system begins to actively intervene in response to the state change, and the vehicle control system plays an active role, but no additional commands are needed for intervention control.

[0116] If 2≤L≤4, F=1, the system begins to issue a risk warning signal, requiring intervention commands for control. Within the power domain, the driver / operator needs to reduce the vehicle speed to within the range of [β2,β3] of the current speed and continue driving, while stabilizing the system's power generation within the range of [β3,1] of the rated operating power. Within the chassis domain, the vehicle stops actively adjusting the suspension system, increases the brake pedal opening and cancels the accelerator pedal opening to meet the speed reduction requirements within the power domain, and maintains the current vehicle steering angle by stabilizing the steering wheel on the current driving trajectory to maintain the current forward state. In the intelligent driving domain, the vehicle expands the preset safety area by β1 times and performs comprehensive system control according to the control requirements of the power domain and chassis domain.

[0117] If F>1, the system approaches an unstable state, requiring strong intervention commands for risk control. Within the power domain, the driver / operator needs to reduce the vehicle speed to within the range [β1, β2] of the current speed. At this point, the system's power generation is reduced to within the rated operating power range [β2, β3]. Within the chassis domain, the vehicle suspension system's active adjustment process transforms into a passive response process. The speed requirements within the power domain are met by adjusting and controlling the brake pedal and canceling the accelerator pedal opening, while simultaneously straightening the steering wheel to maintain stability. After reaching the target speed, the vehicle's steering angle is finely adjusted within the range [-45°, 45°] with the current steering wheel position as a reference, allowing the vehicle to adaptively avoid obstacles. In the realm of intelligent driving, the vehicle expands the preset safety zone by β1 times and performs strong intervention control through adjustments in both the power domain and chassis domain.

[0118] Example 1

[0119] This invention conducts performance feedback mapping research based on state information under unmanned platform domain classification. Under the information fusion and interleaving mechanism, it completes the online intelligent judgment and system decision-making process for platform vehicles through multi-layer fusion and interconnection of underlying state, vehicle judgment, and intelligent decision-making. The following is in conjunction with the appendix... Figure 1 The present invention provides a further explanation of a system intelligent decision-making method for a multi-domain integrated platform.

[0120] For integrated unmanned vehicle chassis, it mainly consists of three parts: power domain, chassis domain and intelligent driving domain. Therefore, in the overall design process, it is necessary to consider the development and design process of multi-domain integrated chassis system, which mainly involves system intelligent decision-making methods and process matching control methods.

[0121] However, since decision-making and control belong to different levels of the vehicle chassis system, this patent mainly involves intelligent decision-making processes for information from different subdomains. In this process, with the increase in sensor and data interaction volume, and the detailed division of different subsystems and subdomains, the amount of real-time state data processed by the vehicle system grows exponentially. Simply stacking state quantities and data dimensions is not the key to solving the problem, nor can it address the challenge of enhancing the autonomy and intelligence of unmanned systems facing complex environments. Therefore, in the process of multi-data fusion, intelligent decision-making methods are needed to solve the problem of system state analysis and determination in the face of overlapping data from multiple domains.

[0122] Step 1: System Status Preprocessing

[0123] 1.1 Data Status Preset

[0124] First, based on the requirements of the multi-domain integrated chassis system, target state variables are selected in different subdomains of the unmanned platform. The selected system variables need to meet the requirements of being able to characterize the performance of the vehicle's modular subsystems.

[0125] In the power domain, parameters related to the characteristics of the vehicle's power system need to be selected for calculation and analysis; in the chassis domain, characteristic parameters related to the suspension state, drive / braking state, and steering subsystem dynamics of the vehicle chassis need to be selected for calculation and analysis; and in the intelligent driving domain, vehicle and external parameters related to multi-vehicle intelligent collaboration need to be selected for comprehensive judgment.

[0126] The specific selection results are as follows: In the power domain, the vehicle speed v, the vehicle power demand P, and the real-time battery SOC are selected as state variables; in the chassis domain, the vehicle vertical displacement d is selected. v Braking force F b Driving force F d Steering angle δ is used as a state variable; within the intelligent driving domain, the maximum longitudinal speed duration interval i is selected. vl Horizontal and longitudinal maximum vehicle speed duration range i vc And the longitudinal distance s from the nearest vehicle l and lateral distance s c As a state variable, where the vehicle speed duration interval i vl and i vc This refers to the longest continuous driving time of a vehicle at its current maximum speed. Together with the distance between the vehicle and its two adjacent vehicles, it represents the safety between the two closest vehicles during multi-vehicle travel.

[0127] 1.2 Multi-system data status preprocessing

[0128] Based on the selected state variables, the specific intervals corresponding to the selected state variables in different domains are first defined, and then the normalization quantization process of different variables is further carried out.

[0129] Within the power domain, the ranges corresponding to vehicle speed v, vehicle power demand P, and current battery SOC are respectively: [v min ,v max ]、[P min ,P max [SOC] min SOC max At this point, normalizing the relevant parameters within the dynamic domain yields the following normalized parameters for different parameters:

[0130]

[0131] Within the chassis domain, the vertical displacement d of the vehicle body v Braking force F b Driving force F d The intervals corresponding to the steering angle δ are respectively: [d v_min ,d v_max ]、[F b_min ,F b_max ]、[F d_min ,F d_max ]、[δ min ,δ max At this point, normalizing the relevant parameters within the chassis domain yields the following normalized parameters for different parameters:

[0132]

[0133]

[0134] Within the intelligent driving domain, the duration of the maximum longitudinal vehicle speed is i vl Maximum lateral and longitudinal speed duration i vc Longitudinal vehicle spacing s l and lateral distance s c The corresponding safe interval ranges are: [i vl_min i vl_max ]、[i vc_min i vc_max ]、[s l_min ,s l_max ]、[s c_min ,s c_max At this point, normalizing the relevant parameters within the intelligent driving domain yields the following normalized parameters for different parameters:

[0135]

[0136]

[0137] Among them, i vl_max and i vc_max These refer to the longest time a vehicle travels at its maximum longitudinal and lateral speeds, respectively. vl_min and i vc_min These refer to the shortest time a vehicle can travel at its minimum longitudinal and lateral speeds, respectively, and are generally taken as 0; however, if the lateral or longitudinal distance between the nearest vehicle and the current vehicle is greater than the safe interval [s], then... l_min ,s l_max ] and [s c_min ,s c_max ], then at this time the parameter N sl and N sc Then the corresponding value is 1.

[0138] Step 2: State Coupling Calculation for Multi-Domain Systems

[0139] Based on the normalized state parameters in different platform subdomains obtained in step one, multi-domain system state coupling calculation is performed. This allows the system parameter determination under the interaction of multiple data dimensions to be transformed into the integrated calculation of system coupling parameters under the dynamic change of parameters. Its main purpose is to solve the problem that the system state is difficult to quantify and intuitively analyze under the influence of multiple parameters and multiple degrees.

[0140] 2.1 Coupled Calculation of Dynamic Domain Parameters

[0141] For N in the dynamic domain v N P and N SOC These three normalized parameters are used to perform parameter coupling calculations.

[0142] Within the power domain, the vehicle's power requirement P is primarily determined by the actual driving speed. Therefore, vehicle speed v and power P are closely related. During the calculation process, it is necessary to avoid redundant power calculations. The actual coupled calculation process requires consideration of N... v and N P The magnitude of N is determined and calculated. Simultaneously, based on the range of battery SOC, N is also... SOC The value is divided into the limiting working interval [0, α]. l ) and adaptable work range [α] l [1], where the working interval boundary value α l Take values ​​within the range of [20%, 25%].

[0143] First, define the exponential critical value α1, where α1 ranges from [0.45, 0.55]. The specific values ​​are as follows:

[0144]

[0145] Further define the power boundary interval value α1, if N v ≥α1, and N P If ≥α1, then the system coupling parameter N1 in the dynamic domain is:

[0146]

[0147] If N v ≥α1, and N P <α1; or N v <α1, and N P If ≥α1, then the system coupling parameter N1 in the dynamic domain is:

[0148]

[0149] If N v <α1, and N P If α1 < α1, then the system coupling parameter N1 in the dynamic domain is:

[0150]

[0151] 2.2 Chassis Domain Parameter Coupling Calculation

[0152] For N in the chassis domain dv N Fb N Fd and N δ These four normalized parameters are used to perform parameter coupling calculations.

[0153] During the calculation, since the driving force and braking force are related, it is necessary to first perform state coupling calculations. A coupling betweenness factor δ is introduced, and the specific coupling parameter N... db The calculation equation is as follows:

[0154]

[0155] In the formula, the coupling intermediate value δ is a nonlinear function, and its calculation formula is as follows:

[0156]

[0157] Based on this, further research on N dv N db and N δThese three parameters are calculated in a coupled manner. Considering that different parameters correspond to the actual state of the vehicle chassis system, the determination of the magnitude of the three numbers is as follows:

[0158]

[0159] In the formula, the function median(N) dv N db N δ ) represents the middle value among these three values.

[0160] The system coupling parameter N2 within the chassis domain is calculated as follows:

[0161]

[0162] 2.3 Coupled Calculation of Parameters in Intelligent Driving Domain

[0163] For N in the intelligent driving domain ivl N ivc N sl and N sc These four normalized parameters are used to perform parameter coupling calculations.

[0164] Within the intelligent driving domain, the longitudinal and lateral distances and actual speeds between adjacent vehicles determine the dynamic safety of multi-vehicle driving. Therefore, the normalized parameters are first determined according to N. ivl N sl and N ivc N sc The calculations are performed in pairs. First, the longitudinal and lateral safety factors N between the two vehicles are calculated separately. pl and N pc Then the calculation formula is as follows:

[0165]

[0166] Next, the parameters are determined, and process parameter N is defined. 3p1 and N 3p2 :

[0167] (1) If N pi =0 (k=l,c), then use N 3p1 =0 and N 3p2 If the result is 0, proceed to the next step of the calculation.

[0168] (2) If N pk =1 (k=l,c), at this time N ivk =N sk Then using N 3p1 =N ivl =N sl and N3p2 =N ivc =N sc Proceed to the next step of the determination and calculation.

[0169] (3) If N pk >1 (k=l,c), at this time N ivk <N sk Then using N 3p1 =N sl and N 3p2 =N sc Proceed to the next step of the determination and calculation.

[0170] (4) If N pk <1 (k=l,c), at this time N ivk >N sk Then we need to calculate N. 3p1 and N 3p2 The corresponding calculation equation at this point is:

[0171]

[0172] If the calculated N 3pj If ≤1 (j=1,2), then N 3pj Take the actual calculated value; if N 3pj If >1 (j=1,2), then N 3pj Take 1.

[0173] N obtained using the above calculation process 3p1 and N 3p2 Furthermore, the system coupling parameter N3 within the intelligent driving domain is calculated. Considering the dynamic distance changes of the vehicle during actual driving, the actual coupling parameter N3 can be obtained as follows:

[0174]

[0175] Step 3: Computational Decision Making under State Coupling

[0176] When the coupled data from each subdomain is aggregated into the integrated controller, the system needs to make overall task decisions based on parameter changes, thereby providing a reference for intelligent vehicle control. Therefore, further research is needed on system decision classification and vehicle control command decisions under classification levels.

[0177] 3.1 Parameter hierarchical processing

[0178] Since the coupling parameters N1, N2, and N3 fall within the interval [0,1], they are divided into four intervals: [0,β1], [β1,β2], [β2,β3], and [β3,1], representing the four processes in the actual decision-making process: system initiation, normal response, control intervention, and risk warning. The values ​​of coupling parameters N1, N2, and N3 are related to priority; larger values ​​indicate higher urgency. The interval boundaries have the following relationship: 0 < β1 < β2 < β3 < 1.

[0179] Based on the interval settings, the value of the interval threshold β0 needs to be determined first. The calculation of β0 mainly considers the difference between the maximum and minimum values ​​of the coupling parameters in different subdomains of the current system. Approaching the critical upper limit value The degree of difference between the maximum value and the critical value. Based on its specific magnitude, the dynamic value of β0 can be determined as follows:

[0180]

[0181] Therefore, based on the value of the interval threshold β0, the corresponding values ​​of β1, β2, and β3 can be calculated as follows:

[0182]

[0183] 3.2 Classification of System Decisions

[0184] Based on the calculated coupling parameters N1, N2, N3, and interval boundary values ​​β1, β2, and β3, a multi-level classification system is determined. Based on the multi-domain coupling data, it is determined whether the corresponding region needs intervention, thereby obtaining the control domain priority L of the system decision. The specific determination process is shown in Table 1.

[0185] Table 1 System Multilevel Classification Process

[0186]

[0187] The system decision priority values ​​in the table are related to the degree of priority. A value of 0 indicates that the system is normal and no action is required. The larger the value, the higher the urgency of the situation.

[0188] Simultaneously, a flag F is set to locate the parameter quantity within the danger warning area, with an initial value of 0. When any one of the coupling parameters N1, N2, and N3 is in the [β3,1] region, the value of F is incremented by 1, and the results are shown in Table 2.

[0189] Table 2 Statistics of System Warning Areas

[0190]

[0191] 3.3 Control Command Decision

[0192] The system's intelligent judgment process is carried out based on the system decision priority L and the number of danger warning areas F.

[0193] If L=0 and F=0, the system will operate passively and normally, without the need for intervention or control via commands.

[0194] If 1≤L≤3, F=0. At this time, the system control system begins to actively intervene in response to state changes, and the vehicle control system plays an active role, but no additional commands are needed for intervention control.

[0195] If 2 ≤ L ≤ 4, F = 1, at which point the system begins to issue a risk warning signal, requiring intervention commands for control. Within the power domain, the driver / operator needs to reduce the vehicle speed to within the range of [β2, β3] of the current speed and continue driving, while stabilizing the system's power generation within the range of [β3, 1] of the rated operating power. Within the chassis domain, the vehicle stops active adjustment of the suspension system, increasing the brake pedal opening and canceling the accelerator pedal opening to meet the speed reduction requirements within the power domain, and maintaining the current vehicle steering angle by stabilizing the steering wheel on the current driving trajectory to maintain the current forward motion. In the intelligent driving domain, the vehicle expands the preset safety zone by a factor of β1, and performs integrated system control according to the operating requirements of the power domain and chassis domain.

[0196] If F > 1, the system approaches an unstable state, requiring strong intervention commands for risk control. Within the power domain, the driver / operator needs to reduce the vehicle speed to within the range [β1, β2] of the current speed. At this point, the system's power generation is reduced to within the rated operating power range [β2, β3]. Within the chassis domain, the active adjustment process of the vehicle suspension system transforms into a passive response process. The speed requirements within the power domain are met by adjusting and controlling the brake pedal and canceling the accelerator pedal opening, while simultaneously straightening the steering wheel to maintain stability. After reaching the target speed, the vehicle's steering angle is finely adjusted within the range [-45°, 45°] (counter-clockwise as the positive direction) based on the current steering wheel position, allowing the vehicle to adaptively avoid obstacles. In the realm of intelligent driving, the vehicle expands the preset safety zone by β1 times and performs strong intervention control through adjustments in both the power domain and chassis domain.

[0197] Example 2

[0198] In the development of intelligent unmanned platform technology, the development of multi-domain information collaboration and fusion decision-making technology based on system performance requirements has become a key factor in platform operation. The purpose of this invention is to provide a system intelligent decision-making method for a multi-domain integrated platform, enabling organic system decision-making during subdomain fusion, information coupling, and dynamic fault changes, thereby achieving decision intervention under the influence of intersecting information.

[0199] The multi-domain vehicle system architecture involved in this invention, such as Figure 1 As shown, the unmanned platform is mainly composed of different subdomains, and the decision-making process mainly involves the research on intelligent decision-making methods and process matching control methods.

[0200] First, this method selects the vehicle speed v, the required power P of the vehicle, and the real-time state of charge (SOC) of the battery in the power domain, and the vertical displacement d of the vehicle body in the chassis domain. v Braking force F b Driving force F d Steering angle δ, and the longitudinal maximum vehicle speed duration range i within the intelligent driving domain. vl Horizontal and longitudinal maximum vehicle speed duration range i vc And the longitudinal distance s from the nearest vehicle l and lateral distance s c These are used as state data within different subdomains. The data is normalized based on the state threshold range to obtain the normalized parameters corresponding to different real-time states.

[0201] Based on this, according to N v N P and N SOC These three normalized parameters are used for system state coupling calculations within the dynamic domain; based on N dv N Fb N Fd and N δ These four normalized parameters are used for system state coupling calculations within the chassis domain; based on N ivl N ivc N sl and N sc These four normalized parameters are used for system state coupling calculations within the intelligent driving domain.

[0202] Therefore, a parameter hierarchical processing and decision classification process based on different subdomain coupling parameters N1, N2 and N3 is further carried out. By determining the control domain priority L and the parameter flag F in the danger warning area, the final control command decision process of the system is completed.

[0203] In summary, this invention belongs to the field of unmanned vehicle swarm control technology, specifically involving a system intelligent decision-making method for a multi-domain integrated platform, comprising three main parts: system state preprocessing, multi-domain system state coupling calculation, and computational decision-making under state coupling. These three parts form the framework and architecture of the system intelligent decision-making, the foundation of which includes vertical multi-dimensional data feature extraction and vertical multi-domain collaboration and parameter coupling. Data feature extraction involves state pre-processing of different subdomains and multi-parameter state coupling calculation, corresponding to the system state preprocessing and multi-domain system state coupling calculation processes. Multi-domain collaboration completes hierarchical processing and decision classification based on multi-domain coupling parameters, synchronously corresponding to the computational decision-making module under state coupling. In this process, the transmission and judgment of system state and the decision-making and issuance of control commands are completed through multi-directional and multi-dimensional data iteration, thereby realizing the integrated processing of system data under integrated geographical and operational conditions, and thus completing the online intelligent decision-making process of the multi-domain platform in the form of a data cluster.

[0204] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A system intelligent decision method of a multi-domain integrated platform, characterized in that, The method includes: Step 1: System state preprocessing; Step 2: Calculation of state coupling in multi-domain systems; Step 3: Computational decision-making under state coupling; Specifically, step 1, system state preprocessing, includes: Step 11: Data status pre-configuration; Step 12: Multi-system data status preprocessing; The data state preset in step 11 is as follows: Within different subdomains of the platform, target state variables are selected, and the selected system variables need to meet the requirements of being able to characterize the performance of the vehicle's modular subsystems. In the power domain, parameters related to the characteristics of the vehicle's power system are selected for calculation and analysis; in the chassis domain, characteristic parameters related to the suspension state, drive / braking state, and steering subsystem dynamics of the vehicle chassis are selected for calculation and analysis; and in the intelligent driving domain, vehicle and external parameters related to multi-vehicle intelligent collaboration are selected for comprehensive judgment. The specific selection result is: select the driving speed within the power domain. v Vehicle power requirements P The battery's real-time SOC is used as a state variable; the vertical displacement of the vehicle body is selected within the chassis domain. d v Braking force F b driving force F d Steering angle δ As a state variable; within the intelligent driving domain, the maximum longitudinal vehicle speed duration range is selected. i vl Maximum speed range in both horizontal and vertical directions i vc And the longitudinal distance to the nearest vehicle. s l and lateral distance s c As a state variable; among which, the vehicle speed duration range i vl and i vc This refers to the longest continuous driving time of a vehicle at its current maximum speed. Together with the distance between the vehicle and the two adjacent vehicles, it represents the safety between the two closest vehicles in a multi-vehicle driving process. In step 2, multi-domain system state coupling calculation is performed based on the normalized state parameters in different platform subdomains obtained in step 1. The multi-domain system state coupling calculation in step 2 specifically includes: Step 21: Coupled calculation of dynamic domain parameters; Step 22: Chassis domain parameter coupling calculation; Step 23: Coupled calculation of intelligent driving domain parameters; Specifically, in step 21, the dynamic domain parameter coupling calculation is as follows: For the dynamic domain N v , N P and N SOC These three normalized state parameters are used to perform parameter coupling calculations within the dynamic domain; Vehicle power within the power domain P Determined by the actual driving speed, therefore vehicle speed v and power P The relationship is closely linked, and the calculation process needs to avoid redundant power calculations; however, the actual coupling calculation process requires consideration of... N v and N P The size is determined and calculated; at the same time, based on the range of battery SOC, it will also be... N SOC The value is divided into the limiting working interval [0, α l ] and adapt to the work area[ α l [1], where the working interval boundary value α l Take values ​​within the range of [20%, 25%]; Define the exponential critical value α 1, α The value of 1 ranges from [0.45, 0.55], and its corresponding values ​​are: Further define the power boundary interval value α 1. If N v ≥ α 1, and N P ≥ α 1. Then, the system coupling parameters within the dynamic domain N 1 is: if N v ≥ α 1, and N P < α 1; or N v < α 1, and N P ≥ α 1. Then, the system coupling parameters within the dynamic domain N 1 is: if N v < α 1, and N P < α 1. Then, the system coupling parameters within the dynamic domain N 1 is: ; Specifically, step 3, which involves computational decision-making under state coupling, includes: Step 31: Parameter hierarchical processing; Step 32: System decision classification; Step 33: Control command decision.

2. The system intelligent decision-making method of the multi-domain integrated platform as described in claim 1, characterized in that, The multi-system data state preprocessing in step 12 is as follows: Based on the selected state variables, the specific intervals corresponding to the selected state variables in different domains are first defined, and the state variables in different subdomains are then normalized and quantized. Within the power domain, driving speed v Vehicle power requirements P The ranges corresponding to the current battery SOC are: [ v min , v max ]、[ P min , P max [SOC] min SOC max At this point, the relevant parameters within the dynamic domain are normalized, and the normalized state parameters corresponding to different parameters are as follows: Vertical displacement of the vehicle body within the chassis domain d v Braking force F b driving force F d and steering angle δ The corresponding interval ranges are: [ d v_min , d v_max ]、[ F b_min , F b_max ]、[ F d_min , F d_max ] 、 [ δ min , δ max At this point, the relevant parameters within the chassis domain are normalized. The normalized state parameters corresponding to different parameters are as follows: Maximum longitudinal speed duration within the intelligent driving domain i vl Maximum speed in both lateral and longitudinal directions i vc Longitudinal vehicle spacing s l and lateral distance s c The corresponding safe interval ranges are: [ i vl_min , i vl_max ]、[ i vc_min , i vc_max ]、[ s l_min , s l_max ]、[ s c_min , s c_max At this point, the relevant parameters within the intelligent driving domain are normalized. The normalized state parameters corresponding to different parameters are as follows: in, i vl_max and i vc_max These refer to the longest time a vehicle can travel at its maximum longitudinal and lateral speeds, respectively. i vl_min and i vc_min These refer to the shortest time a vehicle can travel at its minimum longitudinal and lateral speeds, respectively, and are set to 0; however, if the lateral or longitudinal distance between the nearest vehicle and the current vehicle is greater than the safe range... s l_min , s l_max ]and[ s c_min , s c_max ], then the parameter at this time N sl and N sc Then the corresponding value is 1.

3. The system intelligent decision-making method of the multi-domain integrated platform as described in claim 2, characterized in that, In step 22, the chassis domain parameter coupling calculation is specifically as follows: For the chassis domain N dv , N Fb , N Fd and N δ These four normalized state parameters are used to perform parameter coupling calculations within the chassis domain; During the calculation process, since the driving force and braking force are related, it is first necessary to perform state coupling calculations; the coupling betweenness factor is then introduced. δ Specific coupling parameters N db The calculation equation is as follows: The coupling intermediate value in the formula δ For nonlinear functions, the corresponding calculation formula is: Based on this, further N dv , N db and N δ These three parameters are coupled for calculation; considering that different parameters correspond to the actual state of the vehicle chassis system, the magnitude of the three numbers is determined as follows: In the formula, the function median( N dv , N db , N δ () represents the middle value among these three values; So, the system coupling parameters within the chassis domain N The calculation method is as follows: 。 4. The system intelligent decision-making method of the multi-domain integrated platform as described in claim 3, characterized in that, In step 23, the intelligent driving domain parameter coupling calculation is as follows: Based on the normalized state parameters in the intelligent driving domain N ivl , N ivc , N sl and N sc Perform system state coupling calculations within the intelligent driving domain; Within the intelligent driving domain, the longitudinal and lateral distances and actual speeds between adjacent vehicles determine the dynamic safety of multi-vehicle driving. Therefore, the normalized parameters are first determined according to... N ivl , N sl and N ivc , N sc The calculations are performed in pairs; first, the longitudinal and lateral safety factors between the two vehicles are calculated separately. N pl and N pc Then the calculation formula is as follows: Then, the parameters are determined, and process parameters are defined. N 3p1 and N 3p2 : (1) If N pi =0, k = l , c , then use N 3p1 =0 and N 3p2 =0, proceed to the next step of the determination and calculation; (2) If N pk =1, k = l , c ,at this time N ivk = N sk , then utilize N 3p1 = N ivl = N sl and N 3p2 = N ivc = N sc Proceed to the next step of the determination calculation; (3) If N pk >1, k = l , c ,at this time N ivk < N sk , then utilize N 3p1 = N sl and N 3p2 = N sc Proceed to the next step of the determination calculation; (4) If N pk <1, k = l , c ,at this time N ivk > N sk Then we need to calculate N 3p1 and N 3p2 The corresponding calculation equation at this time is: If the calculation yields N 3pj ≤1, j =1, 2, then N 3pj Take the actual calculated value; if N 3pj >1, then N 3pj Take 1; The results obtained using the above calculation process N 3p1 and N 3p2 Further calculation of system coupling parameters within the intelligent driving domain N 3; Considering the dynamic distance changes of the vehicle during actual driving, the actual coupling parameters can be obtained. N 3 is: 。 5. The system intelligent decision-making method of the multi-domain integrated platform as described in claim 4, characterized in that, In step 31, the coupling parameters obtained from the calculation are... N 1. N 2. N 3. Conduct hierarchical processing of system parameters; Due to coupling parameters N 1. N 2 and N The value 3 belongs to the interval [0, 1], therefore it is divided into [0, 1] and [1]. β 1]、[ β 1, β 2]、[ β 2, β 3] and [ β The four intervals [3, 1] represent the four processes in the actual decision-making process: system startup, normal response, control intervention, and risk warning; among them, the coupling parameters... N 1. N 2 and N The value of 3 is related to the priority; the larger the value, the higher the urgency of the task. The interval boundary values ​​have the following relationship: 0 < β 1< β 2< β 3 < 1; Based on the interval settings, the interval threshold needs to be determined first. β The size of 0; β The calculation of 0 considers the difference between the maximum and minimum values ​​of the coupling parameters in different subdomains of the current system. Approaching the critical upper limit value The degree of difference between the maximum value and the critical value; based on its specific magnitude, it can be determined that... β The dynamic values ​​of 0 are shown below: Then according to the interval threshold β The value of 0 can be calculated β 1. β 2 and β The three values ​​correspond to: ; In step 32, the coupling parameters obtained from the calculation are... N 1. N 2. N 3. Based on the parameter hierarchical processing calculation results, further system decision classification will be carried out; Based on the calculated coupling parameters N 1. N 2. N 3, and interval boundary values β 1. β 2 and β 3. Perform multi-level classification system judgment, and determine whether the corresponding area needs intervention based on multi-domain coupled data, thereby obtaining the control domain priority of system decision. L The specific determination process is shown in Table 1. Table 1 System Multilevel Classification Process The system decision priority values ​​in the table are related to the degree of priority. A value of 0 indicates that the system is normal and no action is required. The larger the value, the higher the urgency of the action. At the same time, set a flag. F To locate the parameter quantity within the danger warning area, its initial value is 0; when the coupling parameter N 1. N 2 and N If any one of the three is in [ β When in the region of 3, 1], F Increment the value by 1, and the results are shown in Table 2. Table 2 Statistics of System Early Warning Areas In step 33, based on the boundary value β 1. β 2. β 3. System Decision Priorities L and number of danger warning areas F Based on the determination results, the system control command decision-making process is initiated; if L =0, F =0, at this time the system operates passively and normally, and no intervention or control is required through commands; If 1≤ L ≤3, F =0, at this time the system control system begins to actively intervene in response to the state change, the vehicle control system plays an active role, but no additional commands are needed for intervention control at this time; If 2≤ L ≤4, F =1, at which point the system begins to issue risk warning signals, requiring intervention commands for control; within the power domain, the driver / operator needs to reduce the vehicle speed to [ ] of the current speed. β 2, β Continue driving within the range of 3] and stabilize the system's power generation within the rated operating power range. β Operating within the range of 3, 1]; within the chassis domain, the vehicle ceases active adjustment of the suspension system, and meets the speed reduction requirements within the power domain by increasing the brake pedal opening and canceling the accelerator pedal opening, while maintaining the current vehicle steering angle by stabilizing the steering wheel on the current driving trajectory, thus maintaining the current forward state; in the field of intelligent driving, the vehicle expands the preset safety zone. β 1x, and perform integrated system control according to the handling requirements of the power domain and chassis domain; if F >1, At this point, the system approaches an unstable state, requiring strong system intervention commands for risk control; within the power domain, the driver / operator needs to reduce the vehicle speed to [ ]. β 1, β 2] Continue driving within the range, at which point the system's power generation capacity is reduced to [ ]% of the rated operating power range. β 2, β 3] Within the operating range; within the chassis domain, the active adjustment process of the vehicle suspension system is transformed into a passive response process, using brake pedal adjustment control and canceling accelerator pedal opening to meet the vehicle speed requirements within the power domain, while simultaneously straightening the steering wheel to maintain stability; after the vehicle speed reaches the target range, using the current steering wheel position as a reference, the vehicle steering angle is finely adjusted within the range of [-45°, 45°] with counterclockwise as the positive direction, allowing the vehicle to adaptively avoid obstacles; in the field of intelligent driving, the vehicle expands the preset safety zone. β It is 1 times larger, and strong intervention control of the system is achieved through adjustments in the power domain and chassis domain.