A method for optimizing the layout of a rail milling and grinding vehicle human-machine interface based on visual perception intensity

Abstract

The application provides a rail milling and grinding vehicle man-machine interface layout optimization method based on visual perception intensity, and belongs to the field of interface ergonomics design of manufacturing equipment. According to the visual perception intensity, the visual perception force grade of the man-machine interaction interface is divided into regions. On this basis, the importance, the viewing frequency and the relevance of the interface parameter information module are comprehensively considered, and a man-machine interface layout optimization model based on visual perception intensity is constructed. The application also provides a rail milling and grinding vehicle man-machine interface layout solving method based on a grey wolf-ant lion intelligent algorithm, which solves the problem that a traditional single heuristic algorithm is easy to fall into local optimization when solving a multi-dimensional domain model. The application has the beneficial effects that: the operation needs of the driver for high comfort and high safety of the rail milling and grinding vehicle man-machine interface interaction are well met, and a very ideal design basis is provided for the optimization design of the man-machine interface of the rail milling and grinding vehicle.

Description

Technical Field

[0001] This invention relates to a method for optimizing the layout of the human-machine interface of a rail milling vehicle based on visual perception intensity, belonging to the field of ergonomic design of manufacturing equipment interfaces. Background Technology

[0002] Currently, railway transportation speeds have greatly increased, and high-speed train operation places higher demands on the smoothness of rails, making equipment for repairing worn rails particularly important. In recent years, my country has introduced rail milling machines to improve railway rail quality; their excellent performance and promising application prospects are noteworthy. As rail maintenance equipment, rail milling machines are undertaking increasingly heavy workloads. To realize the application of rail milling machines in high-precision, high-value-added fields, in-depth and meticulous research on the human-machine interface of rail milling machines is necessary.

[0003] Currently, most rail milling vehicles in China are imported or copied from abroad. Their human-machine interface (HMI) designs focus only on meeting basic functional requirements, neglecting the cognitive characteristics and user experience needs of drivers. However, rail milling vehicle drivers have transitioned from traditional operators to monitors and decision-makers. This necessitates higher cognitive efficiency from HMI information, increasing their cognitive load and workload, leading to low HMI efficiency, operational errors, and safety issues. Therefore, optimizing the HMI design to suit the cognitive characteristics of rail milling vehicle drivers is of significant research and application value for improving HMI efficiency in rail milling vehicles.

[0004] The human-machine interface (HMI) control system of a rail milling vehicle is the control center for rail repair. Rail milling and repair operations are primarily performed by the driver in the cab, and rail repair control is a complex task. Therefore, the degree of interaction between the driver and the HMI directly affects the efficiency and quality of rail repair. Existing HMI optimization research mainly focuses on interface layout, which can meet the basic requirements of human cognitive behavior and has achieved some results. However, considering the complexity of the rail milling vehicle operation process and the multi-dimensional interactive information, especially the high integration and complexity of the control system, existing HMI layout optimization methods are difficult to apply to its multi-dimensional information, multi-process, multi-functional, and multi-object HMI interaction requirements. The main problems to be solved are as follows:

[0005] (1) Existing human-machine interface optimization methods are limited to meeting the basic functional requirements of the system, ignoring the correlation between modules in the interface and the user experience of the driver. They are difficult to meet the complex cognitive requirements of multiple information and multiple processes in the human-machine interaction of rail milling vehicles, which can easily lead to operational errors and untimely monitoring, seriously affecting the quality and safety of rail repair.

[0006] (2) The layout model of the human-machine interface of the rail milling vehicle is designed based on the experience of mechanical engineers to optimize the layout of human cognition and interface functions. It is difficult to systematically take into account the user experience, comfort and safety factors of the driver. It has strong subjectivity and results in low human-machine interaction efficiency.

[0007] (3) When solving the layout optimization model of the rail milling vehicle with multiple decision variables, the traditional single algorithm is more complex in terms of parameter composition and setting, making it difficult to accurately determine the search direction of the optimal solution within the maximum number of iterations, resulting in low solution quality and thus failing to solve the optimal human-machine interface layout scheme of the rail milling vehicle. Summary of the Invention

[0008] This invention aims to provide a method for optimizing the layout of the human-machine interface of a rail milling vehicle based on visual perception intensity, thus providing an effective approach for optimizing the human-machine interaction design of rail milling vehicles.

[0009] To achieve the above objectives, the present invention is accomplished through the following steps:

[0010] S1. Modularization of human-computer interaction interfaces and division of visual perception intensity.

[0011] S2. Determine the usage relationships, operational relationships, and coordination relationships among the various functional modules of the human-computer interaction interface.

[0012] S3. Construct a layout model of the human-machine interface for a rail milling vehicle based on visual perception intensity.

[0013] Furthermore, the human-machine interface layout optimization method for rail milling vehicles based on visual perception intensity is based on the following human-machine interface layout design principles:

[0014] (1) The cognitive needs of drivers should be met, and their experience and cognitive characteristics should be integrated to ensure rapid identification and response to important and critical information. The cognitive patterns of drivers and the importance, viewing frequency and relevance of human-machine interface modules should be fully considered, and the interface layout space should be arranged reasonably.

[0015] (2) The comfort of operation tasks should be met. Each display and control function module should be arranged in a centralized and orderly manner in the comfortable operation area of ​​the personnel according to the viewing frequency, operation sequence and operation correlation, so as to speed up the processing of operation information and effectively improve the efficiency of human-computer interaction and operation comfort.

[0016] (3) Safety requirements should be met, and the impact of safety factors in the driver's cognitive process, operation process and interface design should be fully considered and incorporated into the optimization of the human-machine interface layout as a key layout influencing factor.

[0017] Furthermore, the method for optimizing the layout of the human-machine interface of the rail milling vehicle based on visual perception intensity specifically operates as follows:

[0018] S1. Modularization of the human-computer interaction interface and division of the visual perception area.

[0019] S2. Determine the usage relationships, operational relationships, and coordination relationships among the various functional modules of the human-computer interaction interface.

[0020] S3. Construct the human-machine interface layout model of the rail milling vehicle.

[0021] The aforementioned method for optimizing the layout of the human-machine interface of a rail milling vehicle based on visual perception intensity is based on the following human-machine interface layout design principles:

[0022] (1) The layout of the human-machine interface of the rail milling vehicle should meet the cognitive needs of the driver. In the optimization design of the human-machine interface of the rail milling vehicle, the driver's cognitive process is analyzed in depth first, and the user's work experience and human-machine interaction are combined to ensure the efficiency of task completion. The user experience needs of the driver and the correlation of the human-machine interface modules and the viewing frequency of the modules are fully considered to complete the ergonomic analysis and optimization design of the interface.

[0023] (2) The layout of the human-machine interface (HMI) of the rail milling vehicle should meet the comfort requirements of the monitoring task. During the rail milling operation, due to the intelligent operation of the equipment, the driver needs to closely monitor the dynamic parameter information of the rail milling vehicle and handle any abnormalities in a timely manner. The design of the interface determines the driver's cognitive load and ensures the driver's work comfort. Therefore, when optimizing the HMI layout of the rail milling vehicle, modules with high viewing frequency and high importance should be placed in areas with strong visual perception. This can reduce the driver's cognitive load on the monitoring interface, speed up information processing, and minimize the safety issues caused by untimely information acquisition.

[0024] (3) The layout of the human-machine interface (HMI) of the rail milling vehicle should meet safety requirements. Due to the characteristics of its processing, considering the irreversibility of rail milling and the safety of rail transportation, the rail milling vehicle has extremely high requirements for its operating parameters. During operation, the driver needs to closely monitor the vehicle's working status; even a slight lapse in attention or insufficient response can lead to serious hazards and economic losses. Therefore, the optimized design of the HMI layout for the rail milling vehicle should fully consider the driver's human-machine interaction process, and the interface should be scientifically and rationally designed based on the importance of each module.

[0025] The aforementioned method for optimizing the layout of the human-machine interface of a rail milling machine based on visual perception intensity, wherein step S1, the modularization of the human-machine interface and the division of the visual perception domain, specifically involves the following steps:

[0026] (1) Modular interface: Based on the operation function of the human-machine interface of the rail milling vehicle, the operation status interface is divided into several rectangular regular modules according to the parameter information and display window, without considering the complex shape of the operation status interface and the layout of the area.

[0027] (2) Using the visual analysis module of Jack simulation software, combined with the range of human coronal visual perception area, a driver's visual perception area graphic was drawn. Considering that the driver relies on his eyes to observe and his neck to twist during the monitoring of the milling operation to complete the rail repair, three levels of visual perception intensity areas were divided.

[0028] (3) The driver observes the use of each module. The comfort and stability of each module are related to the visual perception area level and area of ​​each functional module. In order to simplify the calculation process of the module area, the unit grid is processed based on the center line of the working status interface; the visual perception area of ​​the interface module is divided into 5mm as a basic unit.

[0029] The aforementioned method for optimizing the layout of the human-machine interface of a rail milling vehicle based on visual perception regions, wherein step S2 involves determining the usage relationships, operational relationships, and coordination relationships of each functional module of the human-machine interface, specifically using the following method:

[0030] (1) Determining the relational parameters

[0031] Usage relationship refers to the driver's use of various functional modules during the human-machine interface operation of the rail milling vehicle. Usage relationship is determined by the importance of the functional modules and their viewing frequency. Modules with high usage relationships should be arranged as reasonably as possible within the driver's reachable area for quick, comfortable, and safe operation.

[0032] The importance and viewing frequency of interface functional modules are first determined by expert ratings, which assign corresponding importance and usage scores. Then, a group decision method combined with hierarchical analysis is used to transform these scores into importance and viewing frequency indices to represent the importance and viewing frequency of a particular interface functional module. Let the importance rating matrix of the interface functional modules be I, and let a experts rate i modules (i = 16). Then:

[0033] (1) Basic importance coefficient C

[0034] To address the issue that the human-machine interface (HMI) design of rail milling vehicles only meets the rail repair function requirements and neglects the user experience of the driver, this invention starts with expert interviews and questionnaires, uses factor analysis to determine the user experience dimension, employs expert semantic evaluation to assess the correlation between user experience and modules, and finally determines the basic importance C of the HMI of the rail milling vehicle considering user experience based on rough set theory.

[0035] (2) Viewing frequency coefficient F

[0036] Since the human-machine interface of the rail milling vehicle has many parameter information modules and the parameter information is relatively complex, in the analysis of the viewing frequency weight of the 16 modules of the interface, this invention adopts the GEM-AHP optimization method to analyze the viewing frequency coefficient of the milling vehicle interface modules. Based on the original vertical relative weight coefficient, a horizontal scoring matrix of the modules is constructed to obtain the horizontal relative weight, and then the comprehensive weight is obtained.

[0037] The calculation of the viewing frequency weights for each module is as follows:

[0038] Two evaluation indicators, usage frequency and importance, were used to score each component. An evaluation panel of eight experts scored each component using a Likert scale. First, the usage frequency matrix P1 of the target layer was constructed:

[0039] P1=(p ij ) x×y (i=1,2,…,x; j=1,2,…,y) (1)

[0040] In the formula: p ij To evaluate group member i's rating of module j.

[0041] Calculate E1. (E1 is the product of the transpose of the rating matrix P1 of each module and itself):

[0042] E1 = P1 T P1 (2)

[0043] Calculate the eigenvector F1 corresponding to the largest eigenvalue of E1 when it is a single root, and normalize F1 to obtain the relative weight matrix F of the viewing frequency of each interface module.

[0044] The evaluation experts included researchers of the rail milling machine, mechanical engineers, milling machine drivers, and ergonomics professionals. Because the rail milling machine faces complex rail working conditions, its evaluation is inherently subjective and unstable. This invention introduces an expert self-assessment form to correct the expert weighting issues in the analytic hierarchy process (AHP) and group decision-making methods, making the evaluation results more closely reflect reality. The calculation steps are as follows:

[0045] ① Construct an evaluation system for the human-machine interface of the rail milling vehicle's cab;

[0046] ② Experts score themselves using a self-assessment form (as shown in Table 1), and an expert reliability matrix is ​​established. Let the expert group consist of m members, and the evaluation value of the i-th expert be G. i Then G i =a i b i c id i Then the credibility of the i-th expert is

[0047]

[0048] Table 1 Expert Self-Assessment Form

[0049]

[0050] (3) Module correlation coefficient O

[0051] The module correlation degree mainly refers to the degree of functional correlation between two interface modules during operation, that is, whether there is a correlation in monitoring tasks between two interface functional modules during the driver's monitoring interaction with the rail milling vehicle's human-machine interface. Let the correlation matrix be O:

[0052]

[0053] Among them, O id This represents the cooperation relationship between the i-th functional module and the d-th functional module. When constructing the degree of association, [AEIOU] is used to express the strength of the association.

[0054] (4) Constraints

[0055] The constraints of the human-machine interface optimization model for the rail milling vehicle mainly concern the geometric space constraints between the sub-modules to be arranged and the human-machine interface display. The main requirements are: first, to ensure that all modules do not overlap, and that all modules are located within the display screen's area.

[0056]

[0057] Equation (5) indicates that the number of units occupied by the i-th functional module in the visual field of the rail milling car operation status at different levels is equal to the total number of its own units, and the number of units occupied by all functional modules in the visual perception range at all levels is equal to the sum of the areas of all interface modules.

[0058] (5) Human-computer interface layout optimization model based on visual perception intensity

[0059] ① Problem Description:

[0060] When faced with rails in different wear conditions, the adjustment of milling parameters and the complex and ever-changing status of the vehicle and equipment result in varying degrees of importance for the parameter information in the interface modules. Furthermore, for the driver during the monitoring task, the visual perception intensity of parameter information in different areas varies under a standard sitting posture.

[0061] Therefore, the visibility level and area occupied by interface module parameters in different perception intensity areas are important factors affecting driver visual comfort and monitoring efficiency. Based on this, an optimization model for the layout of the human-machine interface of the rail milling vehicle is established to ensure that interface modules with high importance and viewing frequency are placed in areas with high visual perception intensity. The goal of the optimization model is to maximize the visual perception intensity area occupied by the total importance of each interface module.

[0062] ② Assumptions: This invention makes the following assumptions regarding the layout optimization of the human-machine interface for the rail milling vehicle: the size, quantity, and internal composition of all module units of the rail milling vehicle's human-machine interface remain unchanged. There is no overlap between interface modules, and the interface layout area does not exceed the range of the human-machine interface display screen.

[0063] ③ Variable Definition: The main variables of the human-machine interface layout optimization model for rail milling vehicles are:

[0064] W – Total importance, where W = C + O + F

[0065] X — Visual perception intensity levels in different regions, X ij Let be the intensity level of the i-th interface module in the j-th visual perception region.

[0066] Q—The number of grid cells occupied by different levels of visual perception intensity, Q ij This represents the number of grid cells occupied by the i-th module within the j-th visual perception intensity region.

[0067] Z – Overall Strength of Human-Computer Interface Layout Optimization

[0068] S – Total number of human-computer interface units

[0069] ④ Objective function: The human-machine interface layout optimization model for the rail milling vehicle can be expressed as follows:

[0070]

[0071] The present invention provides a method for solving the layout of a rail milling vehicle interface based on a gray wolf-antlion hybrid intelligent algorithm, used in the aforementioned method for optimizing the layout of a rail milling vehicle interface based on a visual perception region. The method includes the following steps:

[0072] Step 1: Initialize the gray wolf population size and hunting stride length and set parameters;

[0073] Step 2: Calculate the fitness value of the gray wolf population, find the globally optimal running position, and select the alpha wolf;

[0074] Step 3: The wolf pack engages in a hunting and running behavior. If a gray wolf discovers a better fitness value during the hunting process, the alpha wolf is replaced.

[0075] Step 4: Determine the critical value. Once the gray wolf pack reaches the critical value, they will switch to pack attack behavior.

[0076] Step 5: Record the current optimal position of the ant and antlion and the global optimal position according to the antlion algorithm formula;

[0077] Step 6: Calculate the fitness value of the antlion and select the first generation of elite antlions;

[0078] Step 7: Randomly roam according to the ant walking formula and let the ant fall into the trap, and solve the ant fitness value;

[0079] Step 8: After the antlion captures an ant, update the antlion's location and select it as an elite antlion;

[0080] Step 9: Determine if the maximum number of iterations has been reached; otherwise, jump to Step 3; if yes, output the global optimal position and end.

[0081] The process and flowchart of the Gray Wolf-Antlion intelligent algorithm are as follows: Figure 6 As shown:

[0082] The aforementioned method for solving the interface layout of a rail milling machine based on the gray wolf-antlion intelligent algorithm, wherein each formula in the solution step is derived from the basic wolf pack algorithm, and its specific calculation principle is as follows:

[0083] (1) Initial Antlion Algorithm

[0084] The antlion algorithm simulates the hunting behavior of antlions in nature. To catch insects, the antlion moves along a circular path, using its abdomen as a plow and its head to absorb loosened particles, digging a cone-shaped pit in the sand. It then hides at the bottom of the pit, leaving its mandibles exposed, to catch insects that slip down. Considering the random movement characteristic of ants while foraging, the simulated ant movement formula is as follows:

[0085] X(t)=[0,cumsum(2r(t1)-1),cumsum(2r(t2)-1),…,cumsum(2r(t) n (7)

[0086] Where cumsum calculates the cumulative sum, n is the maximum number of iterations, t represents the step size of the random walk, and r(t) is a random function, defined as follows:

[0087]

[0088] Where t represents a random walk, and rand is a random number uniformly distributed in the interval [0,1].

[0089] The ant free-running mechanism and random walks are all based on equation (7). At each step of the optimization, the ants update their positions through random walks. To keep the random walks within the search space, they are normalized using the following equation:

[0090]

[0091] Where a i b is the minimum value of the random walk of the i-th variable. i It is the maximum value of the random walk of the i-th variable. It is the minimum value of the i-th variable in the t-th iteration. This represents the maximum value of the variable in the i-th iteration during the t-th iteration.

[0092] When an ant falls into a trap set by an antlion, it will affect the ant's random walk, as expressed by the following equation:

[0093]

[0094]

[0095] Where c t It is the minimum value of all variables in the t-th iteration, d t Let represent a vector containing the maximum value of all variables in the t-th iteration. This indicates the position of the selected j-th ant in the t-th iteration.

[0096] To simulate the antlion's hunting process, it is assumed that the ant is better suited to enter the sand trap than the corresponding antlion and successfully hunts it. Furthermore, the antlion is required to update its location to the latest location of the hunted ant to enhance its search capabilities.

[0097]

[0098] Where t represents the current iteration. This represents the position of the j-th ant selected in the t-th iteration. This represents the position of the i-th ant in the t-th iteration.

[0099] The beneficial effects of this invention are: it effectively meets the driver's needs for a fast, stable, safe, and comfortable operation of the rail milling vehicle interface, and provides an ideal design basis for the human-machine interface design of the rail milling vehicle. Attached Figure Description

[0100] Figure 1 This is a schematic diagram of the interface functional areas in an embodiment of the present invention.

[0101] Figure 2 This is the result of modularizing the functional areas in the embodiments of the present invention.

[0102] Figure 3 This is the modular result of the driver's visual perception intensity of the present invention.

[0103] Figure 4 This is the result of the interface rasterization process in an embodiment of the present invention.

[0104] Figure 5 This is a flowchart of the hybrid intelligent algorithm of the present invention.

[0105] Figure 6 This is a diagram showing the iterative results of the hybrid algorithm in an embodiment of the present invention.

[0106] Figure 7 This is an optimized design scheme according to an embodiment of the present invention.

[0107] Figure 8 This is a verification experiment scenario for an embodiment of the present invention.

[0108] Figure 9 This is a comparison chart of the first fixation time in the verification experiment of the embodiment of the present invention.

[0109] Figure 10(1) and 10(2) These are eye-tracking hotspot comparison diagrams of the optimization effect of the verification experiment in the embodiments of the present invention, wherein Figure 10(1) is the hotspot diagram of the original scheme and Figure 10(2) is the hotspot diagram of the optimized scheme.

[0110] Figure 11(1) and 11(2) These are eye-tracking hotspot comparison diagrams of the optimization effect of the verification experiment in the embodiments of the present invention, wherein Figure 11(1) is the hotspot diagram of the original scheme and Figure 11(2) is the hotspot diagram of the optimized scheme. Detailed Implementation

[0111] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. However, it should be understood that the embodiments are used to explain the present invention and are not intended to limit the present invention.

[0112] This example provides a method for optimizing the layout of the human-machine interface of a rail milling vehicle based on visual perception intensity, including the following steps:

[0113] S1. Modularization of the human-computer interaction interface and division of the visual perception area.

[0114] S2. Determine the usage relationships, operational relationships, and coordination relationships among the various functional modules of the human-computer interaction interface.

[0115] S3. Construct the human-machine interface layout model of the rail milling vehicle.

[0116] The aforementioned method for optimizing the layout of the human-machine interface of a rail milling vehicle based on visual perception intensity is based on the following human-machine interface layout design principles:

[0117] The aforementioned method for optimizing the layout of the human-machine interface of a rail milling machine based on visual perception intensity, wherein step S1, the modularization of the human-machine interface and the division of the visual perception domain, specifically involves the following steps:

[0118] (1) Modular interface: Based on the operation function of the human-machine interface of the rail milling vehicle, the operation status interface is divided into several rectangular regular modules according to the parameter information and display window, without considering the complex shape of the operation status interface and the layout of the area.

[0119] (2) Using the visual analysis module of Jack simulation software, combined with the range of human coronal visual perception area, a driver's visual perception area graphic was drawn. Considering that the driver relies on his eyes to observe and his neck to twist during the monitoring of the milling operation to complete the rail repair, three levels of visual perception intensity areas were divided.

[0120] (3) The driver observes the use of each module. The comfort and stability of each module are related to the visual perception area level and area of ​​each functional module. In order to simplify the calculation process of the module area, the unit grid is processed based on the center line of the working status interface; the visual perception area of ​​the interface module is divided into 5mm as a basic unit.

[0121] See Figure 4 In this embodiment, the modular interface result is overlaid with the visual perception intensity to form the unit gridding result of the human-computer interaction interface in this embodiment. The number of units for each visual perception intensity and the number of units for each module at each visual perception intensity are counted. The statistical results of the number of module units after gridding are shown in Table 2.

[0122] This invention focuses on the operational status monitoring interface of the LINSINGER rail milling machine's milling control system. First, the operational status interface is modularized. Then, the operational status interfaces of each module are merged with the visual perception area and processed into a grid. The number of units within different visual perception intensity levels and the number of cells occupied by each module within each intensity level area are then statistically analyzed. The statistical results of the number of units in each module of the human-machine interface after gridding are shown in Table 2.

[0123] Table 2 Statistical results of the number of units in each module

[0124]

[0125] This invention selected eight experts to conduct expert interviews on the importance, viewing frequency, and module relevance of each functional module within the human-machine interface of a rail milling vehicle, constructing an expert rating matrix for viewing frequency and relevance. To address the impact of the uncertainty of expert weights on the evaluation results, data from six experts were integrated. Experts then rated themselves based on four aspects: professional title, human factors cognition, operational proficiency, and confidence level, determining the expert weights. The expert self-assessment data is shown in Table 3. Group decision-making combined with the analytic hierarchy process (AHP) was used to calculate the viewing frequency weight and relevance coefficient of each interface functional module, as shown in Table 4.

[0126] Table 3. Data from the Expert Self-Assessment Form

[0127]

[0128]

[0129] Table 4 Importance Index and Viewing Frequency Weight Index

[0130]

[0131] Based on the operational characteristics of the LINSINGER rail milling machine, the correlation between the interface functional modules was determined. A score of 1, 3, 5, 7, and 9 was used to represent the correlation between the interface functional modules, with 10 representing the correlation between a module and itself. The closer the matching degree between functional modules, the greater the correlation value. The coordination relationship between the modules of the rail milling machine interface is shown in Table 5.

[0132] Table 5 shows the coordination relationships between interface functional modules.

[0133]

[0134] To verify the effectiveness of the GW-ALO hybrid intelligent algorithm in solving human-computer interface layout models, this invention simultaneously employs the Gray Wolf algorithm, Particle Swarm Optimization algorithm, Genetic Algorithm, Antlion algorithm, and Gray Wolf-Antlion hybrid algorithm for solution quality comparison. The comparison results are as follows: Figure 7 As shown, the solution obtained by the gray wolf algorithm is 276, and the solution obtained by the genetic algorithm is 265. It can be seen that, compared to the other three algorithms, although the gray wolf-antlion hybrid intelligent algorithm does not have a significant advantage in the number of iterations, the quality of the solution is improved by 4.7% compared to the genetic algorithm, 7.2% compared to the particle swarm algorithm, 3.8% compared to the antlion algorithm, and 3.3% compared to the gray wolf algorithm. This demonstrates the superiority of the hybrid intelligent algorithm and its suitability for optimizing the layout of the human-machine interface of a rail milling machine.

[0135] Currently, there is a lot of research on improving and optimizing the Antlion algorithm. Most of the improved algorithms have improved the search speed and search accuracy of the Antlion optimization algorithm to a certain extent, but they have not considered the heuristic information and combined search methods in the iteration process of the Antlion optimization algorithm to improve the ability to find the optimal solution.

[0136] To fully utilize the iterative information in the antlion optimization algorithm, a variable weighting strategy is proposed. Secondly, to prevent the antlion optimization algorithm from getting trapped in local optima, the random running formula of the gray wolf is introduced for alternating iterations, and the ants are allowed to roam using randomly selected antlions, thus escaping local optima. The importance coefficient, viewing frequency coefficient, and relevance coefficient of the interface functional modules, considering user experience, are integrated. Combined with statistical data on the number of units in each region, the algorithm statements are compiled in Matlab based on the gray wolf-antlion hybrid intelligent algorithm flow.

[0137] Assuming a maximum number of iterations of 250 generations and a population size of 150, the optimal layout scheme value obtained after optimization is 278, while the layout scheme before optimization yields a value of 257. The obtained layout schemes are shown in Table 6.

[0138] Table 6 Optimized Interface Layout Scheme

[0139]

[0140]

[0141] The optimized interface layout scheme is as follows: Figure 8 As shown, analysis of the optimized human-machine interface layout reveals that, compared to the original layout, the optimized layout places the key modules requiring close monitoring (including theoretical values ​​of left and right length gauges, cutting speed, current axle value, and vehicle status) within the first and second visual perception domains. Modules with strong relevance and high importance to rail repair operations are more rationally positioned within the driver's more comfortable field of vision, allowing for faster and safer access to parameter information. This facilitates quick and direct monitoring of critical parameters during operation, ensuring safety and aligning with the driver's cognitive characteristics. Next, eye-tracking experiments will be conducted to determine the optimal display method for these parameters.

[0142] According to step S3, the usage relationship parameters, operation relationship parameters and cooperation relationship parameters of the interface function modules are integrated (as shown in Table 6) and substituted into the interface layout model.

[0143] See Figure 5This invention also provides a method for solving the interface layout of a rail milling machine based on a wolf-particle swarm hybrid intelligent algorithm, in order to solve the problem that traditional particle swarm algorithms are prone to getting trapped in local optima when solving multidimensional domain models.

[0144] To achieve the above objectives, the present invention is accomplished through the following steps:

[0145] Step 1: Initialize the gray wolf population size and hunting stride length and set parameters;

[0146] Step 2: Calculate the fitness value of the gray wolf population, find the globally optimal running position, and select the alpha wolf;

[0147] Step 3: The wolf pack engages in a hunting and running behavior. If a gray wolf discovers a better fitness value during the hunting process, the alpha wolf is replaced.

[0148] Step 4: Determine the critical value. Once the gray wolf pack reaches the critical value, they will switch to pack attack behavior.

[0149] Step 5: Record the current optimal position of the ant and antlion and the global optimal position according to the antlion algorithm formula;

[0150] Step 6: Calculate the fitness value of the antlion and select the first generation of elite antlions;

[0151] Step 7: Randomly roam according to the ant walking formula and let the ant fall into the trap, and solve the ant fitness value;

[0152] Step 8: After the antlion captures an ant, update the antlion's location and select it as an elite antlion;

[0153] Step 9: Determine if the maximum number of iterations has been reached; otherwise, jump to Step 3; if yes, output the global optimal position and end.

[0154] Following the steps above, algorithm statements were written in Matlab based on the wolf-particle swarm hybrid intelligent algorithm flow. Assuming a maximum iteration count of 250 generations and a population size of 150 individuals, the optimal layout scheme value was 278 after optimization calculations. The layout scheme value obtained from calculating the unoptimized layout scheme was 257. The obtained layout schemes are shown in Table 7.

[0155] Table 7 Parameters of Human-Machine Interface Functional Modules

[0156]

[0157] Table 8 Optimized Layout Scheme

[0158]

[0159]

[0160] The optimized interface layout scheme is as follows: Figure 8 As shown, analysis of the optimized human-machine interface layout reveals that, compared to the original layout, the optimized layout places the key modules requiring close monitoring (including theoretical values ​​of left and right length gauges, cutting speed, current axle value, and vehicle status) within the first and second visual perception domains. Modules with strong relevance and high importance to rail repair operations are more rationally positioned within the driver's more comfortable field of vision, allowing for faster and safer access to parameter information. This facilitates quick and direct monitoring of critical parameters during operation, ensuring safety and aligning with the driver's cognitive characteristics. Next, eye-tracking experiments will be conducted to determine the optimal display method for these parameters.

[0161] See Figure 8 To demonstrate the effectiveness of the optimized scheme of this invention in improving the efficiency of human-computer interaction and the level of safety and comfort in driver operation, this embodiment uses an eye-tracking experiment for verification. The ErgoLAB human-computer interaction experimental platform used in the experiment consists of a Tobii X2-60 eye tracker (used to capture eye movements and collect data); a signal storage device (for eye calibration and storage of eye movement data before the experiment); and an Acer laptop computer (for simulating the human-computer interface display). This eye tracker uses ErgoLAB software in conjunction with a dongle, and the experimental parameters are set to default settings. Five experienced drivers and five new users with some experience were selected for the experiment, with an average age of 26, and all voluntarily participated. The specific task of the experiment was to monitor the milling and machine parameters of the rail milling machine during the rail milling repair process, following the parameter browsing sequence of the driver's daily operation. This included reading the corresponding parameters, checking whether the parameters met theoretical values ​​and were within a reasonable range, and checking the status information of the milling machine's system.

[0162] See Figure 9 By analyzing and comparing the search time of drivers for a specified region of interest before and after optimization, the extent to which the design schemes before and after optimization meet the needs of drivers for rapid operation can be determined. Table 9 shows the comparison results of the first fixation time before and after optimization.

[0163] Table 9 Comparison of first fixation time before and after optimization.

[0164]

[0165] The comparison results show that the average first fixation time for each area of ​​interest on the interface decreased for both experienced and new users after the optimization. Experienced users saw an average reduction of 8.7 seconds per fixation, while new users experienced a reduction of 2.4 seconds. This indicates that the optimization effectively reduces the time operators spend capturing operational information, thus enabling more efficient task execution. A two-way ANOVA was conducted to analyze the first fixation time before and after optimization, with areas of interest and the optimized interface design as independent variables and first fixation time as the dependent variable. The results of the two-way ANOVA are shown in Table 10.

[0166] Table 10 Results of the two-way ANOVA on first fixation time

[0167]

[0168] Note 1: R-squared = 0.996 (adjusted R-squared = 0.985)

[0169] The F-statistic for the region of interest (ROI) is 2.415, p = 0.000 < 0.05, indicating that the ROI significantly affects first fixation time. The F-statistic for the optimized interface is 11.871, p = 0.021 < 0.05, demonstrating that the optimization significantly impacts first fixation time. This proves that the optimized interface effectively improves the operational efficiency of human-computer interaction compared to the unoptimized interface.

[0170] Fixation duration is an important eye-tracking indicator for measuring the learnability of an interface and the speed of driver cognitive processing. A longer fixation duration indicates a slower reading speed by the driver in that area, leading to difficulty in processing and acquiring information. A longer duration also means that the displayed area of ​​interest contains complex and difficult-to-understand information, resulting in slow information acquisition and a high frequency of viewing by the driver. Conversely, a shorter duration indicates that the information in that area of ​​interest is easy to understand, the information is acquired quickly, and the viewing frequency is low.

[0171] Table 11 Comparison of fixation duration before and after optimization

[0172]

[0173] Table 12 Results of Two-Way Analysis of Variance on Fixation Duration

[0174]

[0175] Note 1: R-squared = 0.996 (adjusted R-squared = 0.985)

[0176] The F-statistic value of the region of interest in the human-computer interface is 56.460, p = 0.003 < 0.05, indicating that the region of interest significantly affects the driver's gaze duration; the F-statistic value of the optimized interface is 13.817, p = 0.021 < 0.05, indicating that the optimization scheme has a significant impact on the gaze duration.

[0177] See Figure 10(1) and 10(2) ,

[0178] From the heat maps before and after optimization, the original and optimized schemes have similar distribution of hot spots and the range of line-of-sight concentration in the core operation area. However, the area of ​​the red hot spots in the original scheme is larger than that in the optimized scheme, indicating that the original scheme has a longer line-of-sight dwell time in the key areas of the monitoring parameters, and the gaze time is also relatively longer than that of the circular cross section.

[0179] The trajectory diagrams before and after optimization show that the number of fixation points decreased from 38 to 28. Furthermore, the fixation points in the optimized diagram are more concentrated within the strongest visual perception range (I). Compared to the original trajectory diagram, the optimized diagram demonstrates more logical line-of-sight, less line-of-sight intersection, and a shorter visual fixation distance. This further proves that the optimized solution better meets the driver's needs for comfort, safety, and efficiency.

[0180] Experimental results show that the interface design scheme produced by the visual field-based human-machine interface layout optimization design model for rail milling vehicles improves the efficiency, comfort, and safety of driver operations during monitoring tasks compared to the original scheme. The interface design generally meets the needs of operational comfort and safety. Furthermore, through comparison of first gaze time before and after optimization and two-way ANOVA, it can be confirmed that the optimized human-machine interface effectively improves operational comfort and safety without sacrificing human-machine interaction efficiency. In fact, the optimized interface can shorten operation time to varying degrees, thereby improving human-machine interaction efficiency.

[0181] The above description represents preferred embodiments of the present invention and is not intended to limit the invention. Those skilled in the art can still modify the above technical solutions or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for optimizing the layout of the human-machine interface of a rail milling vehicle based on visual perception intensity. S1. Modularization and visual perception area division of the human-computer interaction interface. S2. Determine the basic importance, viewing frequency coefficient, and correlation coefficient of each functional module in the human-computer interaction interface. Determining the basic importance coefficient C: Starting from expert interviews and questionnaires, factor analysis was used to determine the user experience dimension. Expert semantic evaluation was used to assess the correlation between user experience and modules. Based on rough set theory, the basic importance coefficient C of the rail milling car human-machine interface considering user experience was finally determined. Determining the viewing frequency coefficient F: In the viewing frequency weight analysis of the 16 modules of the interface, the group decision-making method-analytic hierarchy process (AHP) optimization method was used to analyze the viewing frequency coefficient of the milling machine interface modules. Based on the original vertical relative weight coefficients, a horizontal scoring matrix of the modules was constructed to obtain the horizontal relative weights, and then the comprehensive weights were derived. Two evaluation indicators, usage frequency and importance, were used to score each component. An evaluation panel of eight experts scored each component using a Likert scale. First, the usage frequency matrix P1 of the target layer was constructed. P1=(p ni ) x×y (n=1,2,…,x;i=1,2,…,y)(1) In the formula: p ni To evaluate group member n's rating of module i, Calculate E1, which is the product of the transpose of the rating matrix P1 for each module and P1 itself: E1=P1 T P1(2) Calculate the eigenvector F1 corresponding to the largest eigenvalue of E1 when it is a single root, and normalize F1 to obtain the relative weight matrix F of the viewing frequency of each interface module. Introduce an expert self-evaluation table to correct the expert weight problem in the analytic hierarchy process and group decision method. Experts conduct self-evaluation based on the self-evaluation table to establish an expert reliability matrix. Determine the module correlation coefficient O: The module correlation coefficient represents the degree of functional correlation between two interface modules during operation, that is, whether there is a correlation in monitoring tasks between two interface functional modules during the driver's monitoring interaction with the rail milling vehicle's human-machine interface. Let the correlation matrix be O: ⑶ in, O id This represents the cooperation relationship between the i-th functional module and the d-th functional module; S3. Construct the layout model of the human-machine interface for the rail milling vehicle, as follows: Assuming that the size, quantity, and internal composition of all modules in the human-machine interface of the rail milling vehicle remain unchanged, The constraints are: ensure that all modules do not overlap and that all modules are located within the display screen area. (4) Where Q represents the number of grid cells occupied by different levels of visual perception intensity. ij Let S be the number of grid cells occupied by the i-th module in the j-th visual perception intensity region, and let S be the total number of human-computer interface cells. Equation (4) indicates that the number of units occupied by the i-th functional module in the visual perception area of ​​the rail milling car is equal to the total number of its own units in different levels of visual perception area, and the number of units occupied by all functional modules in all levels of visual perception area is equal to the sum of the areas of all interface modules. The human-machine interface layout optimization model of the rail milling car is represented as follows: (5) Where W represents the total importance, W = C + O + F; X represents the visual perception intensity level of different regions. ij Let Z be the intensity level of the i-th interface module in the j-th visual perception region, and Z be the comprehensive intensity of human-computer interface layout optimization.

2. The method for optimizing the layout of the human-machine interface of a rail milling vehicle based on visual perception intensity according to claim 1, characterized in that, The modularization and visual perception area division of the human-computer interaction interface in step S1 are specifically carried out as follows: (1) Modular interface: Based on the operation function of the human-machine interface of the rail milling vehicle, the operation status interface is divided into several rectangular regular modules according to the parameter information and display window, without considering the complex shape of the operation status interface and the layout of the area. (2) Using the visual analysis module of Jack simulation software, combined with the range of human coronal visual perception area, a driver's visual perception area graphic was drawn. Considering that the driver relies on his eyes to observe and his neck to twist during the monitoring of the milling operation to complete the rail repair, three levels of visual perception intensity areas were divided. (3) The unit gridding process is carried out based on the center line of the work status interface; the visual perception area of ​​the interface module is divided into 5mm as a basic unit.

3. The method for optimizing the layout of the human-machine interface of a rail milling vehicle based on visual perception intensity according to claim 1, characterized in that, The interface layout of the rail milling machine is solved using a hybrid intelligent algorithm based on gray wolf-antlion, specifically including the following steps: Step 1: Initialize the gray wolf population size and hunting stride length and set parameters; Step 2: Calculate the fitness value of the gray wolf population, find the globally optimal running position, and select the alpha wolf; Step 3: The wolf pack engages in a hunting and running behavior. If a gray wolf discovers a better fitness value during the hunting process, the alpha wolf is replaced. Step 4: Determine the critical value. Once the gray wolf pack reaches the critical value, they will switch to pack attack behavior. Step 5: Record the current optimal position of the ant and antlion and the global optimal position according to the antlion algorithm formula; Step 6: Calculate the fitness value of the antlion and select the first generation of elite antlions; Step 7: Randomly roam according to the ant walking formula and let the ant fall into the trap, and solve the ant fitness value; Step 8: After the antlion captures an ant, update the antlion's location and select it as an elite antlion; Step 9: Determine if the maximum number of iterations has been reached; otherwise, jump to Step 3; if yes, output the global optimal position and end.

4. The method for optimizing the layout of the human-machine interface of a rail milling vehicle based on visual perception intensity according to claim 3, characterized in that, The specific operational principle is as follows: (1) Initial Antlion Algorithm The antlion algorithm simulates the hunting behavior of antlions in nature. To catch insects, the antlion moves along a circular path, using its abdomen as a plow and its head to absorb loosened particles, digging a cone-shaped pit in the sand. It then hides at the bottom of the pit, leaving its mandibles exposed, to catch insects that slip down. Considering the random movement characteristic of ants while foraging, the simulated ant movement formula is as follows: X(t)=[0,cumsum(2r(t1)-1),cumsum(2r(t2)-1),…,cumsum(2r(t n )-1),] (6) Where cumsum is the cumulative sum, n is the maximum number of iterations, t represents the step size of the random walk, and r(t) is a random function defined as follows: (7) Where t represents a random walk, and rand is a random number uniformly distributed in the interval [0,1]. The ant free-running mechanism and random walks are all based on equation (6). In each step of the optimization, the ants update their positions by random walking. In order to keep the random walks within the search space, they are normalized using the following equation: (8) Where a i b is the minimum value of the random walk of the i-th variable. i It is the maximum value of the random walk of the i-th variable. It is the minimum value of the i-th variable in the t-th iteration. This represents the maximum value of the variable in the i-th iteration at the t-th iteration. When an ant falls into a trap set by an antlion, it will affect the ant's random walk, as expressed by the following equation: (9) (10) Where c t It is the minimum value of all variables in the t-th iteration, d t Let represent a vector containing the maximum value of all variables in the t-th iteration. This indicates the position of the selected j-th ant in the t-th iteration. To simulate the antlion's hunting process, it is assumed that the ant is better suited to enter the sand trap than the corresponding antlion to successfully hunt it. Furthermore, the antlion is required to update its location to the latest location of the hunted ant to enhance its search capabilities. (11) Where t represents the current iteration. This represents the position of the j-th ant selected in the t-th iteration. This represents the position of the i-th ant in the t-th iteration.

Images