Wire harness distribution method based on regional centralized architecture
By adopting a harness distribution method based on a regional centralized architecture, and utilizing clustering algorithms and fuzzy inference to optimize harness layout, the problems of code redundancy and insufficient communication capabilities in distributed architecture are solved, achieving harness optimization and cost reduction.
Patent Information
- Application Number
- CN202410888908.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-04
- Publication Date
- 2024-11-01
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
In existing technologies, distributed electronic and electrical architectures suffer from code redundancy and insufficient communication capabilities, making it difficult to meet the communication and computing needs of intelligent vehicles. There is also a lack of intelligent design methods, especially insufficient research on the close integration of the characteristics and requirements of autonomous driving function algorithms.
A harness distribution method based on a regional centralized architecture is adopted. The communication topology is designed using a clustering algorithm. Combined with particle swarm optimization and FAR fuzzy inference, multiple communication topology layout optimization schemes are generated, and the optimal harness layout is finally determined.
It optimizes the length and cost of wire harnesses, reduces wire harness weight and expenses, and provides an automated design algorithm to quickly determine the optimization scheme, which is applicable to practical engineering improvements.
Smart Images

Figure CN118886142A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of automotive electronic technology, and in particular to a wiring harness distribution method based on a regional centralized architecture. Background Art
[0002] As vehicles become more intelligent and connected, the shortcomings and limitations of the traditional automotive electronic and electrical architecture, the distributed electronic and electrical architecture, are gradually exposed. This distributed electronic and electrical architecture has a large code redundancy and its communication capabilities can no longer meet the increasing demand for ECU (Electronic Control Unit) components.
[0003] The new generation of electronic and electrical architecture that is gradually developing is a centralized electronic and electrical architecture based on domain controllers and Ethernet backbone communication networks. This architecture can improve the limitations of traditional electronic and electrical architectures and vehicle networks and meet the communication and computing needs of smart cars. At present, based on this centralized electronic and electrical architecture, the design of its communication topology layout is mostly an engineering design case, such as the X-EEA 3.0 architecture installed on the G9 model of Xiaopeng Motors. It reduces the number of ECUs and thus simplifies the wiring harness topology. However, it is difficult to achieve comprehensive optimization based on experience and intuition, and the design efficiency needs to be improved. In terms of academics, related research is mainly focused on the engineering development of distributed electronic and electrical architectures based on design tools such as Preevision. There is little research on multi-domain communication topology design, and there is a lack of exploration of intelligent design methods. In addition, the research on related design modeling and methods is not closely integrated with the characteristics and requirements of autonomous driving function algorithms. In the field of vehicle design, there are some intelligent design methods, but they are mainly focused on mechanical structure-related designs. There has been no research specifically exploring intelligent design methods for autonomous driving electronic and electrical systems, and research in this area is still missing.
[0004] Therefore, a wiring harness distribution method based on a regional centralized architecture has become an urgent problem to be solved. Summary of the invention
[0005] The technical problem to be solved by the present invention is to provide a wiring harness distribution method based on a regional centralized architecture, introduce a clustering algorithm to design the communication topology relationship, use an intelligent design method to generate multiple communication topology layout optimization schemes, and ultimately only decide on the optimal scheme.
[0006] To solve the above technical problems, the present invention provides a technical solution: a wiring harness distribution method based on a regional centralized architecture, comprising the following steps:
[0007] S1. Analyze and model the characteristics of regional centralized electronic and electrical architecture;
[0008] S2, automatically design the topological connection relationship between each regional gateway and the underlying components of the system based on the clustering algorithm;
[0009] S3. Taking the topological connection relationship as a constraint, multi-objective optimization design is performed on the layout of regional gateways and computing platforms for low cost and lightweight, and multiple Pareto optimal solutions are generated;
[0010] S4. Based on FAR fuzzy reasoning, intelligent reasoning and decision making are performed on alternative options to automatically select the best wiring harness layout solution.
[0011] Furthermore, under the regional centralized architecture of step S1, the system components include a computing platform, a regional gateway, and underlying components.
[0012] Furthermore, the vehicle geometric coordinate system should be established before modeling in step S1, and the specific method is as follows:
[0013] A vehicle geometric coordinate system oxyz is established, where x, y, and z represent the coordinate axes in the length, width, and height directions of the vehicle respectively. The origin o is located at the geometric center of the vehicle in the top view in the x and y directions, and its height in the z direction is the same as the lowest point of the vehicle tire.
[0014] Furthermore, in step S2, the Kmeans clustering algorithm is used to divide the lower-level components into k clusters, each cluster is assigned a regional gateway; the Kmeans clustering algorithm is used to solve the topological connection relationship between the lower-level components and the regional gateways; the lower-level components are all selected to be connected to the regional gateways with the same color.
[0015] Furthermore, in step S3, the layout positions of each computing platform and each regional gateway are designed using a particle swarm algorithm:
[0016] The mass per unit length of the bus is denoted by ρ(b i ), the price per unit length is recorded as c(b i );
[0017] Optimization variables:
[0018]
[0019] Objective function:
[0020] J(X)=ω1J1(X)+ω2J2(X)
[0021]
[0022] Among them, ||e i ||1 represents e i The L1 norm of , i.e., the Manhattan distance;
[0023] Constraints:
[0024]
[0025] J1(X)≤W max
[0026] J2(X)≤C max
[0027] Among them, J1(X) is the harness weight function and J2(X) is the harness cost function.
[0028] Further, in step S4, the FAR fuzzy reasoning model is used to make inference decisions on key parameters; let ω1 = 1, ω2 = ω obj k obj , where k obj To unify the scaling factor of the two objective functions, ω obj is the weight coefficient of the two objective functions after the order of magnitude is unified; the method of reasoning and decision making based on the FAR fuzzy reasoning method is as follows:
[0029]
[0030] Defining decision variables Select the Reichenbach fuzzy implication type and use the triangle and trapezoidal membership functions to obtain the fuzzy inference control system:
[0031]
[0032]
[0033] Design the inference controller u as:
[0034]
[0035] Under the guidance of control u, the fuzzy inference control system is guided to Because And μ F (·) is a bijective function, and the decision variable Finally stabilized at According to the steady-state value of the decision variable Determine the weight parameter value This parameter value can be used for optimization.
[0036] The advantages of the present invention over the prior art are that the present invention connects the underlying components, regional gateways and computing platforms in a Manhattan distance manner. Although this Manhattan distance connection method cannot accurately represent the direction of the actual wiring harness, it also approximately simulates the wiring method of the actual wiring harness along the beams, grooves, ribs and the like in the horizontal, longitudinal and vertical directions of the vehicle to a certain extent. In practical applications, this method can be used as an automated design algorithm to quickly determine a set of optimization solutions, and the designer can perform more detailed engineering implementation and improvement of the solution based on actual engineering experience and design constraints. In addition, in the actual wiring process, some underlying components can share part of the wiring harness, without having each underlying component have an independent line connected to the regional gateway. BRIEF DESCRIPTION OF THE DRAWINGS
[0037] Figure 1 It is a schematic diagram of mapping the underlying components of a wiring harness distribution method based on a regional centralized architecture of the present invention onto a vehicle layout.
[0038] Figure 2 It is a schematic diagram of topological connection relationships solved by using a Kmeans clustering algorithm in a wiring harness distribution method based on a regional centralized architecture of the present invention.
[0039] Figure 3 It is a schematic diagram of the optimal layout of regional gateways and computing platforms of a wiring harness distribution method based on a regional centralized architecture of the present invention.
[0040] Figure 4 It is a schematic diagram of the overall process of optimizing the wiring harness distribution of an electronic and electrical architecture according to a wiring harness distribution method based on a regional centralized architecture of the present invention.
[0041] Figure 5 The present invention is a flowchart of the operation of a Kmeans clustering algorithm of a wire harness distribution method based on a regional centralized architecture.
[0042] Figure 6 The present invention is a flowchart of the operation of a particle swarm algorithm for a wire harness distribution method based on a regional centralized architecture.
[0043] Figure 7 This is a diagram of the wiring harness cost and weight optimization results of a wiring harness distribution method based on a regional centralized architecture of the present invention. DETAILED DESCRIPTION
[0044] Various exemplary embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be noted that unless otherwise specifically stated, the relative arrangement of components and steps, numerical expressions and numerical values set forth in these embodiments do not limit the scope of the present invention.
[0045] At the same time, it should be understood that for the convenience of description, the sizes of the various parts shown in the drawings are not drawn according to the actual proportional relationship.
[0046] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the invention, its application, or uses.
[0047] Technologies, methods, and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods, and equipment should be considered as part of the specification.
[0048] In all examples shown and discussed herein, any specific values should be interpreted as merely exemplary and not as limiting. Therefore, other examples of the exemplary embodiments may have different values.
[0049] It should be noted that like reference numerals and letters refer to similar items in the following figures, and therefore, once an item is defined in one figure, it need not be further discussed in subsequent figures.
[0050] The following is a further detailed description of a wiring harness distribution method based on a regional centralized architecture of the present invention in conjunction with the accompanying drawings.
[0051] Working principle of the present invention:
[0052] Combined with Figure 1-7 The specific implementation process of the wiring harness distribution method based on the regional centralized architecture of the present invention is as follows:
[0053] The present invention provides a wiring harness layout method for reducing wiring harness length, cost and weight based on a regional centralized electrical and electronic architecture, comprising the following steps:
[0054] (1) Under the regional centralized architecture, the system components mainly include three types of components: computing platform, regional gateway, and underlying components (sensors, ECU, actuators). The vehicle geometric coordinate system oxyz, x, y, z respectively represent the coordinate axes in the length, width, and height directions of the vehicle. The origin o is located at the geometric center of the vehicle in the top view in the x and y directions, and the height in the z direction is the same as the lowest point of the vehicle tire. In the present invention, the position information of the underlying components is known. The Kmeans clustering algorithm is used to divide the lower-level components into k clusters, and each cluster is assigned a regional gateway. Take a vehicle with a length of 4.5m, a width of 2m, a height of 1.8m, and 90 underlying sensors, actuators, and ECU components as an example. The distribution of its components is shown in the attached figure. Figure 1 shown.
[0055] The Kmeans clustering algorithm is used to solve the topological connection relationship between the above 90 underlying components and regional gateways. The number of regional gateways is set to 4. The clustering results are shown in the attached figure. Figure 2 As shown. Among them, the four square marks are the regional gateways after convergence, and their position coordinates are (1.26, 0.53, 0.68), (1.28, -0.71, 0.63), (-1.48, 0.53, 0.58), (-1.46, -0.69, 0.60). The bottom components (circular marks) are all connected to the regional gateways with the same color. The black diamond mark is the initial position of the autonomous driving computing platform (domain controller) (the three-dimensional coordinates are (0.20, 0, 1.00)).
[0056] (2) Based on the above results, with the goal of low cost and lightweight system topology, an optimization model is built. Taking the particle swarm algorithm as an example, the layout positions of each computing platform and each regional gateway are designed. The unit length mass of the bus is denoted as ρ(b i ), the price per unit length is recorded as c(b i ).
[0057] Optimization variables:
[0058]
[0059] Objective function:
[0060] J(X)=ω1J1(X)+ω2J2(X)
[0061]
[0062] Among them, ||e i ||1 represents e i The L1 norm of , that is, the Manhattan distance.
[0063] Constraints:
[0064]
[0065] J1(X)≤W max
[0066] J2(X)≤C max
[0067] (3) Based on the FAR fuzzy reasoning model, the key parameters are inferred and decided. Let ω1 = 1, ω2 = ω obj k obj , where k obj To unify the scaling factor of the two objective functions, ω obj is the weight coefficient of the two objective functions after they are unified in magnitude, and reasoning and decision making need to be performed based on the FAR fuzzy reasoning method.
[0068]
[0069] Defining decision variables Select the Reichenbach fuzzy implication type and use the triangle and trapezoidal membership functions to obtain the fuzzy inference control system:
[0070]
[0071]
[0072] Design the inference controller u as:
[0073]
[0074] Under the guidance of control u, the fuzzy inference control system is guided to Because And μ F (·) is a bijective function, and the decision variable Finally stabilized at According to the steady-state value of the decision variable Determine the weight parameter value This parameter value can be used for optimization.
[0075] Based on the solution of Kmeans clustering algorithm in (1), the cost control level j1 is set to 6 and the weight control level j2 is set to 10, that is, the focus is on reducing the weight of the wiring harness. Based on the optimization method and FAR fuzzy inference system, the decision variables can be obtained The steady-state solution The weight parameter is Solve for k obj =0.0151. So J(X)=J1(X)+ω obj k obj J1(X)=J1(X)+0.151J1(X). Solving the optimization model, we can get the optimal layout positions of the four regional gateways as (1.20, 0.41, 0.45), (1.20, -0.60, 0.45), (-1.60, 0.41, 0.45), (-1.60, -0.60, 0.45), and the optimal layout position of the computing platform is (-0.24, -0.07, 0.45). The optimal layout effect of the regional gateway and computing platform is shown in the attached figure. Figure 3 .
[0076] The present invention connects the underlying components, regional gateways and computing platforms in a Manhattan distance manner. Although this Manhattan distance connection method cannot accurately represent the direction of the actual wiring harness, it also approximately simulates the wiring method of the actual wiring harness along the beams, grooves, ribs and other wiring in the horizontal, longitudinal and vertical directions of the vehicle to a certain extent. In practical applications, this method can be used as an automated design algorithm to quickly determine a set of optimization schemes, and designers can perform more detailed engineering implementation and improvement of the scheme based on actual engineering experience and design constraints. In addition, in the actual wiring process, some underlying components can share part of the wiring harness, without having to make each underlying component have an independent line connected to the regional gateway. Although the model used in this article does not consider this harness sharing situation, it can be proved by the following brief analysis: if the optimization model proposed in this section is used without considering the harness sharing factor to reduce the system harness length, then in practice, after considering the harness sharing factor, the optimal layout obtained by the optimization model can still reduce the actual harness length.
[0077] In the above example, the variation curves of J(X), J1(X) and J2(X) are shown in the attached figure. Figure 7 As shown in the figure, the harness weight function J1(X) dropped by 7.91kg from the initial value of 9.08kg, a decrease of 12.90%; the harness cost function J2(X) dropped from the initial value of 646.60 yuan to 548.16 yuan, a decrease of 15.22%. Among them, 8.98kg and 631 yuan were calculated based on the regional centralized architecture after clustering and before optimization.
[0078] The present invention and its embodiments are described above, and such description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if ordinary technicians in the field are inspired by it, without departing from the purpose of the invention, they can design a structure and embodiment similar to the technical solution without creativity, which should belong to the protection scope of the present invention.
Claims
1. A wiring harness distribution method based on a regional centralized architecture, characterized in that: The following steps are included: S1. Analyze and model the characteristics of regional centralized electronic and electrical architecture; S2, automatically design the topological connection relationship between each regional gateway and the underlying components of the system based on the clustering algorithm; S3. Taking the topological connection relationship as a constraint, multi-objective optimization design is performed on the layout of regional gateways and computing platforms for low cost and lightweight, and multiple Pareto optimal solutions are generated; S4. Based on FAR fuzzy reasoning, intelligent reasoning and decision making are performed on alternative options to automatically select the best wiring harness layout solution.
2. The wiring harness distribution method based on regional centralized architecture according to claim 1, characterized in that: In the regional centralized architecture of step S1, the system components include a computing platform, a regional gateway, and underlying components.
3. The wiring harness distribution method based on regional centralized architecture according to claim 2, characterized in that: Before modeling in step S1, the vehicle geometric coordinate system should be established. The specific method is as follows: Establish the vehicle geometric coordinate system oxyz, where,, represent the coordinate axes in the length, width, and height directions of the vehicle respectively. The origin in the and directions is located at the geometric center of the vehicle in the top view, and the height in the and directions is the same as the lowest point of the vehicle tire.
4. The wiring harness distribution method based on regional centralized architecture according to claim 3, characterized in that: In step S2, the Kmeans clustering algorithm is used to divide the lower-level components into k clusters, each cluster is assigned a regional gateway; the Kmeans clustering algorithm is used to solve the topological connection relationship between the lower-level components and the regional gateways; the lower-level components are all selected to be connected to the regional gateways with the same color.
5. The wiring harness distribution method based on regional centralized architecture according to claim 4, characterized in that: In step S3, the particle swarm algorithm is used to design the layout positions of each computing platform and each regional gateway: The mass per unit length of the bus is denoted by ρ(b i ), the price per unit length is recorded as c(b i ); Optimization variables: Objective function: J(X)=ω1J1(X)+ω2J2(X) Among them, ||e i ||1 represents e i The L1 norm of , i.e., the Manhattan distance; Constraints: J1(X)≤W max J2(X)≤C max Among them, J1(X) is the harness weight function and J2(X) is the harness cost function.
6. The wiring harness distribution method based on regional centralized architecture according to claim 5, characterized in that: In step S4, the FAR fuzzy reasoning model is used to make inference decisions on key parameters; let ω1 = 1, ω2 = ω obj k obj , where k obj is the scaling factor to unify the magnitude of the two objective functions, ωobj is the weight coefficient of the two objective functions after the magnitude is unified; the method of reasoning and decision making based on the FAR fuzzy reasoning method is as follows: Defining decision variables Select the Reichenbach fuzzy implication type and use the triangle and trapezoidal membership functions to obtain the fuzzy inference control system: Design the inference controller u as: Under the guidance of control u, the fuzzy inference control system is guided to Because And μ F (·) is a bijective function, and the decision variable Finally stabilized at According to the steady-state value of the decision variable Determine the weight parameter value This parameter value can be used for optimization and solution.
