128 results about "Configuration space" patented technology

The invention relates to the technical field of high-speed rail fault detection, and discloses an intelligent fault detection system for a high-speed rail. A data acquisition module of the system acquires a heterogeneous operation log data set, including vibration, temperature, current time sequence data and control signal data, of a high-speed rail core component in a preset monitoring period; the stability evaluation module performs multi-dimensional stability analysis on the data set to generate vibration, temperature and current stability indexes and signal integrity indexes; the detection configuration optimization module takes a multi-dimensional index as an index, performs self-adaptive search in a pre-constructed high-dimensional fault detection configuration space, and determines a dynamic detection step length; and the fault detection execution module drives the multi-sensor array to execute periodic detection based on the dynamic detection step length, generates a real-time monitoring data stream, and extracts abnormal features through a fault mode recognition engine. The system can comprehensively capture the operation state of the component, adaptively adjust the detection configuration, improve the fault detection accuracy and real-time performance, and guarantee the operation safety of the high-speed rail.

Embodiments of the present disclosure provide a method and apparatus for reporting asset information, a non-transitory computer readable storage medium, and an electronic device. The method for reporting asset information includes: reading, in a case where a target device supports memory expansion, Serial Presence Detect (SPD) information of a designated memory bank mounted to the target device from a configuration space register of a Memory Expander Controller (MXC) of the target device by means of a Basic Input Output System (BIOS), wherein the target device is a device that supports a Computer Express Link (CXL) protocol, which is an open interconnection standard; and reporting, in a case where target SPD information has been read, the target SPD information to a Central Processing Unit (CPU) of the target device as system asset information by means of the BIOS.

The invention relates to the technical field of chip packaging, in particular to a PCIe (peripheral component interface express) equipment function number border crossing prevention method supporting ARI (automatic repeat interface). The hardware layer protection process is as follows: S1-1, hot plugging; s1-2, if it is detected that the equipment declarates to support ARI, the PCIe controller hardware circuit immediately sets an equipment number register to zero; s1-3, only allowing the PCIe controller of the system to read the configuration space unidirectionally during the low level period of the reset signal PERST #; the interception process of the driving layer is as follows: S2-1, in an equipment enumeration stage, an operating system drives to dynamically read an ARI capability register; s2-2, if the equipment declarates to support ARI, forcing an equipment number register to be equal to 0, and releasing an original equipment number bit space for function number expansion; s2-3, border crossing access is blocked; and S2-4, synchronizing the state register. Compared with the prior art, correctness of device function numbers is guaranteed through two dimensions of hardware register state solidification and driver layer access truncation, and'perception-free 'fault tolerance of the PCIe device under ARI / traditional mode switching is achieved for the first time.

Provided in the embodiments of the present disclosure are a chip management network configuration method, a chip management network system, a computing device, a storage medium and a computer program product, which are applied to a target chip. The target chip comprises a plurality of different types of chiplets, each chiplet containing one or more modules and being configured with a management network configuration serial number, and the management network configuration serial number being configured on the basis of the type of the chiplet and its encapsulation position in the target chip. The chip management network configuration method comprises: acquiring architecture information of the target chip, at least comprising the management network configuration serial numbers configured for the chiplets; on the basis of the architecture information of the target chip, dividing a preset address space into a plurality of sub-configuration spaces; and, on the basis of the management network configuration serial numbers, allocating corresponding sub-configuration spaces among the plurality of sub-configuration spaces to the target chip and each chiplet in the target chip, wherein the sub-configuration spaces allocated to the modules contained in each chiplet are different.

The invention belongs to the technical field of robot path planning, and particularly discloses a robot path planning method based on large language model guidance, and the method comprises the steps: discretizing an environment space, and building a configuration space of a robot and an obstacle; calculating a collision area in real time and segmenting a free space into connected components; generating a prompt text; calling a large language model to analyze the text and outputting a recommended obstacle list; constructing a path tree; generating a new node; generating candidate paths based on new node backtracking, and allowing node poses to shift in a preset range through a tolerance verification mechanism; and when the target area is connected with the connected component, adding the path into the solution set library, and selecting an optimal solution according to a weighted cost function of a minimum path length and a total obstacle moving distance. The search strategy is optimized by using the semantic understanding ability of the large language model, the efficiency and adaptability of path planning in a dynamic environment are remarkably improved, and the method is suitable for scenes such as warehouse logistics and disaster rescue.

The invention discloses a reinforcement learning aided evolutionary algorithm for parameter optimization in algorithm configuration, which comprises the following steps of: performing parameterized definition on four core operators, coding all adjustable parameters, and constructing a unified parameter configuration space; then randomly initializing a configuration population, and inputting the configuration population into an optimization framework assisted by reinforcement learning; the framework is internally provided with an intelligent agent and is executed in each generation of iteration: operating an evolutionary algorithm according to parameter configuration and evaluating performance; the intelligent agent adaptively selects a predefined alignment strategy based on a population state, and guides mutation operation to generate a filial generation; performing environment selection and population updating based on performance evaluation; and finally, calculating rewards according to population performance, and training and updating the agent decision model. And after iteration to a termination condition, outputting configuration with optimal performance and a corresponding algorithm thereof. According to the method, an evolutionary algorithm search space is constructed through comprehensive parameterization, a reinforcement learning agent is introduced to adaptively process variable-length coding, and a double-layer optimization framework is adopted to efficiently search optimal algorithm configuration.

The invention belongs to the field of hoisting track planning, and particularly discloses a bridge girder erection machine double-hook collaborative hoisting track planning method considering space-time coupling constraint, which comprises the following steps: constructing a kinematic model of a bridge girder erection machine and a three-dimensional obstacle geometric model of a hoisting operation environment, and forming a multi-dimensional configuration space; generating a hoisting operation sequence according to the type attribute of the to-be-hoisted component; in the global path planning stage, a sampling-based global path planning algorithm is adopted to carry out search sampling in a multi-dimensional configuration space, and a double-hook kinematics cooperative constraint detection mechanism is implanted to obtain an initial collision-free path; in the local trajectory optimization stage, a trajectory smoothing model based on quadratic programming is constructed, and a smooth and synchronous time state trajectory is generated by minimizing jerk and introducing synchronization error hard constraint. The problems of space-time coordination and anti-swing control in double-hook hoisting operation are effectively solved, and the optimal balance of efficiency and safety is achieved.

The invention provides a PXE network card initialization method and device compatible with 64-bit addressing and a medium, and relates to the technical field of network card initialization, and the method comprises the steps: discovering a network card device; reading a PCI configuration space of the network card equipment, obtaining information of an extended ROM base address register, and mapping a network card Option ROM to a memory shadow region; when the access of the memory mapping input and output mode fails, the initialization code performs read-write operation on the hardware register through a self-defined capability structure in a PCI configuration space of the network card equipment; based on the successfully accessed hardware register, network card hardware initialization is completed, and a PXE protocol stack structural body is established in a memory; registering system firmware as PXE starting equipment, and entering a protocol interaction stage with a PXE server; according to the invention, smooth development of a subsequent network boot flow is guaranteed, and the deployment success rate and the system stability in a large-scale batch installation scene are improved.

The invention belongs to the technical field of automatic sand blasting, and provides a mechanical arm sand blasting track generation method and system, a medium and a robot. The mechanical arm sand blasting track generation method comprises the steps that key process parameters of a sand blasting system are discretized in a layered mode; calculating the actual angle of each joint of the positioner; coordinates of spraying points are calculated; the joint angle of the mechanical arm is calculated, and the collision-free effective posture of the mechanical arm meeting the sand blasting process requirement is screened; a collision-free and process-feasible configuration space diagram is constructed, the vertex of the configuration space diagram is composed of all spraying points, and the edge of the configuration space diagram is composed of paths of the adjacent spraying points; and according to the collision-free and process-feasible configuration space diagram, an optimal path from a preset starting vertex to a target vertex is searched for and serves as a mechanical arm sand blasting track for optimizing the kinematics performance and the sand blasting process quality at the same time. The automatic level of sand blasting treatment, the consistency of treatment quality and the overall operation efficiency can be improved.

The invention discloses a micro-grid multi-objective optimization configuration method and system based on double-layer modeling, and relates to the technical field of micro-grid planning optimization, and the method comprises the following steps: obtaining equipment operation parameters, branch connection relations and typical daily multi-source power time sequence data of a micro-grid; constructing a first configuration space and constructing a configuration layer model according to the first configuration space; generating a physical topology based on the configuration layer model and performing time sequence simulation to obtain a power flow path reconstruction sequence; extracting loss time sequence distribution of each branch and constructing a loss migration feature set; constructing a branch loss fluctuation constraint based on the loss migration feature set, and feeding back the branch loss fluctuation constraint to the configuration layer model as a feasibility constraint to obtain a corrected configuration layer model; the optimal configuration scheme meeting the feasibility constraint is searched in the first configuration space by adopting the multi-objective optimization algorithm, so that the problem of frequent reconstruction of a power flow path caused by dynamic change of power injection of a distributed power supply and an energy storage device along with time in the operation process of the micro-grid is solved.

The invention provides a straddle-type vehicle which can obtain a handlebar-mounted air bag which can easily ensure the configuration space of an air bag unit and can easily ensure the rigidity of a handlebar. The straddle-type vehicle is provided with: a head pipe (18); a handlebar (21) attached to the head tube (18); and an airbag unit (61) capable of deploying an airbag (65), in which the handlebar (21) is provided with: left and right handlebar tube sections (21b) that expand to the outside in the vehicle width direction above the head tube (18); and a horizontal tube (21c) that bridges the left and right handlebar tube sections (21b), and at least a portion of the airbag unit (61) is disposed in a space between the left and right handlebar tube sections (21b) and the horizontal tube (21c).

The application discloses a kind of building passive energy-saving strategy generation method and system, it is related to passive energy-saving strategy generation technical field, including compiling and archiving strategy generation task book, solidification target caliber source, target hierarchical rule and global partition facade roof four layers hard constraint gate and conflict record field;According to task book, it is divided into strategy unit according to orientation shelter, enclosure structure, opening restriction, operation constraint and construction window consistency and solidifies account snapshot, generates evidence chain at unit granularity and executes sufficiency gate and benchmark backout;Establish configuration space and eliminate configuration that triggers hard constraint;Finally generate strategy package and screen sorting, output strategy combination, implement in stages and verify read-write results.The method disclosed in the application realizes strategy explainable, implementable and verifiable by target caliber and hard constraint gate solidification, evidence chain backout, mutually exclusive dependency alternative constraint and three-stage screening, reduces conflict rework and improves energy-saving and budget controllability.

A first wireless device may obtain a configuration of a three-dimensional (3D) space. The 3D space may include a set of discrete points. The configuration may include a set of established compressed beamforming (CBF) reports. Each of the set of established CBF reports may be associated with at least one discrete point of the set of discrete points. The first wireless device may receive an updated CBF report from a second wireless device. The first wireless device may calculate a location of the second wireless device based on the updated CBF report and the set of established CBF reports. The first wireless device may be an access point (AP), a base station, a network node, a network entity, or a transmission reception point (TRP). The second wireless device may be a station (STA) or a UE.

The application discloses a virtual PCI device for multi-domain isolated communication and an implementation method thereof, which comprises the following steps: dividing the memory mapping IO address space of the virtual PCI device into a general configuration space and a private configuration space, storing a first type of register for generally configuring a plurality of data channels in the general configuration space, and storing a second type of register for specially configuring a single data channel in the private configuration space. The protocol limitation of the virtual PCI device can be broken, and a plurality of data channels can be served by one virtual PCI device.

The application provides a disk formatting method and terminal, wherein key data is stored in a preset initial configuration space, so that a boot program can directly find the key data in the initial configuration space and does not need to enter an operating system, thereby achieving that the key data of the terminal can be acquired in a boot stage when the terminal is formatted in batches. In addition, before the disk is formatted, the key data is written into a memory for backup, so that the key data is prevented from being lost during formatting, and the key data of the terminal is ensured to be still stored in the disk after the formatting operation.

This invention discloses an automated tuning method and system for large language model service configurations, belonging to the fields of artificial intelligence and cloud computing technology. The invention includes an offline analysis module for performance profiling of the large language model (LLM) service system under a finite configuration set and for collecting training data; an online search module for exploring and evaluating the performance of different configurations in the configuration space based on a Monte Carlo tree search algorithm; and a simulator module for receiving the configurations generated by the online search module and predicting the performance metrics of the configurations under a target workload based on a logical step simulation model. This invention effectively solves the problems of low search efficiency and poor performance of existing automatic tuning methods in existing large language model service frameworks (such as vLLM, SGLang, etc.) due to the huge number of configuration parameters (reaching a petabyte-scale search space) and the complex dependencies and mutual exclusion relationships between parameters.

The invention relates to a high-precision assembly path planning method for an artificial intelligence cooperative industrial robot, and the method comprises the steps: replacing the path planning of a high-dimensional configuration space of an assembly system with the multi-group path planning of a low-dimensional configuration space of an industrial robot under the condition that the mutual collision of the industrial robots is not considered; and then iteratively searching and detecting conflicts among the candidate paths, and eliminating the conflicts among the candidate paths by utilizing low-cost candidate path replacement and new candidate path generation so as to obtain collision-free candidate paths among industrial robots, and combining the collision-free candidate paths to obtain a plurality of groups of collision-free assembly system path planning results. And selecting the path planning result with the optimal cost from the path planning results of the assembly system as the final path planning result of the assembly system. Through the design, dimension reduction in the path planning process is achieved, the path planning calculation efficiency is improved, and real-time path planning of a complex industrial robot assembly system is better met.

This invention proposes a robotic arm obstacle avoidance path planning method based on an improved ant colony algorithm, relating to the fields of robot path planning and intelligent optimization algorithms. This method endows the joint configuration space with a non-uniform metric structure induced by a metric tensor G(q), where G(q) is derived from the normalized joint inertia matrix W. I The algorithm consists of three parts: the singularity gradient outer product term and the obstacle spacing gradient outer product term. Local geodesic distances are approximated using the mean of the metric tensors at both ends of the node. All edge weights are pre-calculated and cached during the PRM graph construction phase. The ant colony uses the reciprocal of the geodesic distance as a heuristic function, drives non-uniform pheromone evaporation using the normalized value of the metric tensor trace increment, and uses the weighted sum of geodesic length cost and inertia-weighted velocity mutation penalty as the comprehensive path cost. The beetle whisker algorithm performs pre-search and completes non-uniform pheromone initialization under geodesic metrics. This method effectively improves path safety, continuity, and dynamic adaptability.

Disclosed are a server boot and operation method and apparatus, a server, and a storage medium. The server boot and operation method includes: during a power-on self-test (POST) process of a server, loading an Option ROM of each target peripheral device according to a peripheral device loading sequence and by using a target configuration space; in response to detecting that the target configuration space is insufficient, detecting remaining spaces of other modules in a preset address space; and in response to determining that the remaining spaces meet a required space corresponding to the target configuration space, reallocating a preset address space to ensure that the reallocated configuration space for the peripheral device loading module is greater than or equal to a sum of the target configuration space and the required space.

The application relates to the technical field of AI education interaction, and discloses an AI education interaction system based on motion capture. The system comprises four modules, namely a motion data acquisition module, a motion quality analysis module, a teaching parameter configuration module and an interaction result generation module. The motion data acquisition module acquires a historical motion monitoring data set of a target user in a preset historical time period through a motion sensor; the motion quality analysis module traverses the data set, analyzes and determines a motion stability factor and a motion accuracy factor; the teaching parameter configuration module indexes the two factors, optimizes search in a teaching configuration space, and determines a target teaching parameter; and the interaction result generation module carries out interaction with the user according to a preset teaching item based on the target teaching parameter, acquires a real-time interaction data set, and obtains a teaching target interaction result after AI recognizer identification. The system can accurately monitor motion, scientifically analyze quality and provide personalized guidance, and improve the teaching experience.

The invention provides a mechanical arm path planning method and system, and the method comprises the steps: taking an initial collision-free path of a mechanical arm as a center corridor, carrying out the boundary expansion of an obstacle region in a configuration space, and constructing a collision-free convex polyhedron sequence in the center corridor, determining a safe passage channel from an initial configuration to a target configuration through the convex polyhedron sequence; determining a dynamic constraint condition and a motion smoothness index of the mechanical arm, and performing constraint optimization on the configuration track of the mechanical arm based on the dynamic constraint condition and the motion smoothness index in the constraint boundary of the safe passage to obtain a joint constraint track of the mechanical arm in the configuration space; and the joint space track is issued to a mechanical arm joint controller of the robot to be executed, and when it is sensed that a new unmodeled obstacle invades the safe passing channel, a new path planning request of the robot is triggered. On the basis of the scheme, mechanical arm track optimization under the dynamic safety channel constraint can be achieved.

The application belongs to the field of path planning, and particularly relates to a robot formation planning method based on a five-dimensional configuration space. The method comprises the following steps: workspace and configuration space definition: the workspace is a two-dimensional plane for robot movement, and the configuration space is a five-dimensional space containing formation centroid, direction and scaling factor; sampling method: random sampling in the configuration space, and ensuring that the sampling points do not cause collision between robots; distance function: defining the distance between two points in the configuration space, and eliminating the dependence on the number of robots; search: searching in the configuration space using an algorithm, and constructing a collision-free path tree; path generation: converting the path in the configuration space into the path in the robot workspace, and smoothing using a spline curve. The robot formation can pass through a narrow channel, and split and merge when encountering a large obstacle. The application provides a high-efficiency, flexible and collision-free global path planning method for a multi-robot system.

The invention relates to the technical field of tool path planning, and discloses a combined curved surface constant residual height fixed-axis machining tool path generation method and a storage medium, and the method comprises the steps: firstly, discretizing a combined curved surface into a triangular mesh, and calculating the height of a collision-free tool location point above a machining part through adaptive quadtree sampling; forming a net-shaped structure of a discrete three-dimensional configuration space; projecting the processing area to the space for cutting, and determining an effective processing area; taking the boundary of the area as an initial seed polygon, outwards expanding line segments of the boundary into a cylindrical surface in a limited range, intersecting the cylindrical surface with the configuration space to generate a next circle of equidistant cutter location points, connecting the equidistant cutter location points to form a new cutter path, and iterating the process until the processing area is filled. According to the method, tool path selfing is fundamentally eradicated, the method is completely independent of the normal direction of the curved surface, the method is not sensitive to the model quality, the interference-free and equal-residue-height high-quality tool path can be generated in a full-automatic and high-robustness mode, and the reliability, efficiency and quality of complex curved surface numerical control programming are greatly improved.

This invention belongs to the field of robot path planning technology, specifically disclosing a robot path planning method guided by a large language model. The method includes: discretizing the environmental space to establish a configuration space for the robot and obstacles; calculating the collision region in real time and segmenting the free space into connected components; generating prompt text; calling the large language model to parse the text and output a recommended obstacle list; constructing a path tree; generating new nodes; backtracking to generate candidate paths based on the new nodes, and allowing node poses to shift within a preset range through a tolerance verification mechanism; when the target region is connected to a connected component, adding the path to the solution set library, and selecting the optimal solution based on a weighted cost function that minimizes the path length and the total distance moved by the obstacles. This invention utilizes the semantic understanding capability of the large language model to optimize the search strategy, significantly improving the efficiency and adaptability of path planning in dynamic environments, and is applicable to scenarios such as warehousing logistics and disaster relief.

The application belongs to the field of hoisting trajectory planning, and particularly discloses a method for planning a double-hook coordinated hoisting trajectory of a bridge girder erection machine considering space-time coupling constraints, which comprises the following steps: constructing a kinematic model of the bridge girder erection machine and a three-dimensional obstacle geometric model of a hoisting operation environment to form a multi-dimensional configuration space; generating a hoisting operation sequence according to the type attribute of a component to be hoisted; in the global path planning stage, a global path planning algorithm based on sampling is used to search and sample in the multi-dimensional configuration space, and a double-hook kinematic coordination constraint detection mechanism is implanted to obtain an initial collision-free path; in the local trajectory optimization stage, a trajectory smoothing model based on quadratic programming is constructed, a smooth and synchronous time-state trajectory is generated by minimizing jerk and introducing a synchronization error hard constraint. The application effectively solves the space-time coordination and anti-swing control problems in the double-hook hoisting operation, and realizes the optimal balance between efficiency and safety.

The application provides a PXE network card initialization method, device and medium compatible with 64-bit addressing, relates to the technical field of network card initialization, and comprises the following steps: discovering a network card device; reading the PCI configuration space of the network card device, acquiring extended ROM base address register information, and mapping the network card Option ROM to a memory shadow area; when memory mapping input / output mode access fails, the initialization code performs read / write operation on the hardware register through a self-defined capability structure in the PCI configuration space of the network card device; based on the successfully accessed hardware register, network card hardware initialization is completed, and a PXE protocol stack structure body is established in the memory; the system firmware is registered as a PXE starting device, and a protocol interaction stage with a PXE server is entered; and the application guarantees the smooth development of a subsequent network booting process, and improves the deployment success rate and system stability in a large-scale batch deployment environment.

The application provides a multi-index dynamic priority unmanned aerial vehicle cooperative path planning method, and the path planning method comprises the following steps: acquiring robot parameter information, path constraint information, task environment information and algorithm parameter information; acquiring obstacle information in a simulation environment, and constructing a simulation environment grid map, so as to uniformly divide a map into grid maps with equal proportions; obtaining an unmanned aerial vehicle cooperative path planning configuration space according to the simulation environment grid map; and searching in the unmanned aerial vehicle cooperative path planning configuration space by using an unmanned aerial vehicle cooperative path planning algorithm based on a WM-MI-CS algorithm, so as to obtain a feasible unmanned aerial vehicle cooperative path. According to an example task requirement, the multi-index dynamic priority unmanned aerial vehicle cooperative path planning method has the advantages of low operation cost, short task completion time and fast algorithm speed.

Costs associated with configurations corresponding to a maneuver type(s) may be stored in a transition state(s) volume. The same memory volume may be used for storing cost values that correspond different maneuver types and different vertices in a graph of a configuration space. In at least one embodiment, to share a memory volume between maneuver types, the system may determine a cost for a machine to reach a configuration of a configuration space using various different maneuver types. The system may then evaluate one or more of the costs to determine which of the costs to store at one or more memory location(s) corresponding to the configuration (e.g., a point in a memory volume). Cost values for the memory volume may be efficiently determined using kernel-style processing.

Method for motion planning of an autonomous motor vehicle (10), wherein the method comprises the following steps: - Determining (S1) an initial state of the motor vehicle (10) and an environment of the motor vehicle (10) by a sensor device (12) of the motor vehicle (10); - Determining (S2) a target position and a target state of the motor vehicle (10) in the environment by a computing device (14) of the motor vehicle (10); - Determining (S3) configuration samples (20, 20', 20'') in a configuration space (18) by the computing device (14), wherein the configuration samples (20, 20', 20'') are generated depending on a motion model of the motor vehicle (10) and thus impermissible movements of the motor vehicle (10) are excluded, wherein the configuration space (18) represents the environment of the motor vehicle (10) and the configuration samples (20, 20',20'') possible positions of the motor vehicle (10) with a respective corresponding state of the motor vehicle (10), wherein the configuration samples (20, 20', 20'') are determined using a sampling-based algorithm in predefined sampling steps, wherein in each sampling step a plurality of configuration samples (20, 20', 20'') are determined in parallel using at least one graphics processing unit (15) of the computing device (14), wherein the configuration samples (20, 20', 20'') generated in each sampling step are based on the state and position of the configuration samples (20, 20', 20'') of the preceding sampling step, wherein respective sequences of configuration samples (20, 20', 20'') possible paths (22,22') of the motor vehicle (10) through the configuration space (18) to reach the target position and target state;- wherein the configuration space (18) is rasterized into several grid cells (26) (S4) and a grid cell object distance (28) is determined from a center of a respective grid cell (26) to one of the nearest objects (24) in the environment of the respective grid cell (26), wherein a motor vehicle shape of the motor vehicle (10) in the configuration space (18) is modeled as a motor vehicle circle model comprising a circle (30) or a composition of several circles, wherein for a respective grid cell (26) a distance of the motor vehicle (10) to the nearest object (24) in the environment is calculated by a difference of the grid cell object distance (28) and a radius of a circle (30) of the motor vehicle circle model,- evaluating (S5) the possible paths (22, 22') and / or newly generated Configuration sample values ​​(20, 20',20'') after each sampling step by the computing device (14) depending on a path criterion, wherein the path criterion checks at least the calculated distance of the motor vehicle (10) to the objects (24) in the environment for the presence of at least a minimum distance, excluding paths and / or configuration sampling values ​​for which the minimum distance is not present;- providing (S6) at least one possible path (22') as a motion trajectory of the motor vehicle to a control device (16) of the motor vehicle (10) for controlling the autonomous motor vehicle (10) to the target position in the target state;- wherein objects (24) in the environment are classified as low and high objects (24), wherein for high objects (24) the presence of the minimum distance is checked, wherein for low objects (24) the configuration space (18) is subdivided into a height profile grid with slope information, wherein for configuration sampling values ​​(20, 20', 20''),which are located near a low object (24), a wheel position of the motor vehicle is determined from the state of the respective configuration scan values ​​(20, 20', 20'') in relation to the low object (24), wherein, depending on the wheel position, a lateral distance of a respective wheel (30) to the low object (24) is determined, whereby the possible path (22) and / or the configuration scan value for which the lateral distance is below a wheel distance threshold value is excluded.

Method for generating an optimized neural network architecture taking into account the hardware constraints of a hardware target. The present invention relates to a method for generating an optimized neural network architecture taking into account the hardware constraints of a hardware target (20), the method comprising: obtaining an algorithmic task (100); defining a neural search space (104) as a function of the algorithmic task (100); defining a hardware configuration space (102) of the hardware target; searching for a neural network architecture for the hardware target using a genetic algorithm (108) to optimize evaluation criteria (106), the evaluation criteria (106) being at least one hardware performance and at least one algorithmic performance;and evaluate the hardware performance(s) by using a performance predictor if the performance predictor has a reliability above a reliability threshold or the hardware performance(s) being determined by a performance evaluation by running the neural network on the hardware target (20) if the performance predictor has a reliability below the reliability threshold. Figure for the abstract: 2;