A truck configuration intelligent recommendation method, system, device and medium

Abstract

The present application relates to the technical field of vehicle configuration recommendation, and discloses a truck configuration intelligent recommendation method, system, device and medium. The method comprises the following steps: obtaining daily operation conditions and corresponding static configuration data of a plurality of sold trucks; extracting the internal relationship between different configuration parameter combinations in the static configuration data of each sold truck and the mapping relationship between the daily operation conditions and truck performance indexes; determining the rationality evaluation score of different configuration parameter combinations according to the internal relationship between the different configuration parameter combinations, and obtaining the matching relationship between the daily operation conditions of each sold truck and the corresponding static configuration data in combination with the mapping relationship between the daily operation conditions and the truck performance indexes; and constructing a configuration condition matching model through the matching relationship between the daily operation conditions and the corresponding static configuration data of all sold trucks, so as to perform truck configuration intelligent recommendation, and the recommendation accuracy and efficiency are both higher.

Description

Technical Field

[0001] This invention relates to the field of vehicle configuration recommendation technology, and in particular to a method, system, device and medium for intelligent truck configuration recommendation. Background Technology

[0002] Trucks, as a key representative of the commercial vehicle sector, differ from passenger cars in that they are primarily used for commercial road transportation operations. Because trucks offer a greater variety of cargo types, weights, routes, and operating conditions, customers have far more options for vehicle configurations when purchasing them compared to passenger cars. Configurations that match the transportation scenarios and operating conditions can lead to higher economic efficiency, ensure the vehicle operates stably and continuously in good technical condition, effectively improve customer satisfaction and loyalty, and promote repeat purchases. Conversely, if the recommended configuration does not match the operating conditions, it can result in poor vehicle fuel economy, or even frequent breakdowns, leading to customer complaints, decreased satisfaction, and even customer churn.

[0003] Currently, truck marketing and configuration recommendations in the commercial vehicle industry mainly rely on the personal experience of sales personnel. Newcomers to the industry often find it difficult to effectively and accurately recommend matching vehicle configurations to potential customers, resulting in low marketing communication efficiency. Customer complaints and grievances due to mismatches between recommended configurations and actual working conditions are frequent occurrences. Summary of the Invention

[0004] The purpose of this invention is to provide a method, system, device, and medium for intelligent truck configuration recommendation, which can solve the problems of low accuracy and efficiency in configuration recommendation.

[0005] To address the aforementioned technical problems, embodiments of the present invention provide a truck configuration intelligent recommendation method, comprising the following steps: Obtain daily operating conditions and corresponding static configuration data for several sold trucks; The inherent relationship between different combinations of configuration parameters in the static configuration data of each sold truck is extracted by random forest model, and the mapping relationship between the daily operating conditions and truck performance indicators of each sold truck is extracted by gradient boosting decision tree model. Based on the inherent relationship between different configuration parameter combinations, the rationality evaluation score of different configuration parameter combinations is determined. The rationality evaluation score of each configuration parameter combination of each sold truck and the mapping relationship between daily operating conditions and truck performance indicators are input into the gradient boosting decision tree model to obtain the matching relationship between the daily operating conditions of each sold truck and the corresponding static configuration data. A configuration condition matching model is constructed by matching the daily operating conditions of all sold trucks with the corresponding static configuration data. By configuring a working condition matching model, intelligent recommendations are made for the corresponding static configuration data of the user's truck based on the actual daily operating conditions of the truck.

[0006] Optionally, the step of constructing a configuration condition matching model by matching the daily operating conditions of all sold trucks with the corresponding static configuration data includes: Based on the mapping relationship between the daily operating conditions and truck performance indicators of each sold truck extracted by the gradient boosting decision tree model, the performance indicators of the truck under each daily operating condition are predicted. Based on the degree of deviation between the predicted performance indicators of the truck under each daily operating condition and the actual performance indicators of the truck under the corresponding daily operating condition, the weight of the matching relationship between the daily operating conditions of each sold truck and the corresponding static configuration data is determined. A configuration condition matching model is constructed by matching the daily operating conditions of sold trucks with different weights with the corresponding static configuration data.

[0007] Optionally, the step of constructing a configuration condition matching model by matching the daily operating conditions of all sold trucks with the corresponding static configuration data includes: Based on the truck's R&D and design data, the matching relationship between the daily operating conditions of all sold trucks and the corresponding static configuration data is optimized; among which, the R&D and design data includes the engineering boundary design to ensure the normal operation of the truck. A configuration condition matching model is constructed by matching the optimized daily operating conditions of all sold trucks with the corresponding static configuration data.

[0008] Optionally, the step of constructing a configuration condition matching model by matching the daily operating conditions of all sold trucks with the corresponding static configuration data includes: Based on customer feedback on configuration combination optimization suggestions from successfully verified configuration combination cases in historical sales orders, the matching relationship between the daily operating conditions and corresponding static configuration data of all sold trucks is optimized. A configuration condition matching model is constructed by matching the optimized daily operating conditions of all sold trucks with the corresponding static configuration data.

[0009] Optionally, the step of intelligently recommending corresponding static configuration data for user trucks based on the actual daily operating conditions of the user's trucks through a configuration working condition matching model includes: By configuring a working condition matching model, the system intelligently recommends static configuration data for multiple trucks based on the actual daily operating conditions of the user's trucks. Based on the user's truck purchase needs, the weights of the three truck purchase objectives—economy, reliability, and purchase cost—are adjusted. Then, based on these adjusted weights, static configuration data that meets the truck purchase needs is selected from multiple intelligently recommended static configuration data.

[0010] Optionally, the static configuration data includes the following configuration parameters: fuel type, application type, vehicle series, horsepower, version, edition, transmission model, rear axle ratio, tire model, and drive type.

[0011] Optionally, the daily operating conditions are determined by the following truck operating data: road grade, altitude, gradient, load status, ambient temperature, and average vehicle speed.

[0012] Embodiments of the present invention also provide a truck configuration intelligent recommendation system, comprising: The data acquisition module is used to acquire the daily operating conditions and corresponding static configuration data of several sold trucks; The feature extraction module is used to extract the inherent relationship between different combinations of configuration parameters in the static configuration data of each sold truck through a random forest model, and to extract the mapping relationship between the daily operating conditions and truck performance indicators of each sold truck through a gradient boosting decision tree model. The matching relationship acquisition module is used to determine the reasonable evaluation score of different configuration parameter combinations based on the inherent relationship between different configuration parameter combinations, and input the evaluation score of each configuration parameter combination of each sold truck and the mapping relationship between daily operating conditions and truck performance indicators into the gradient boosting decision tree model to obtain the matching relationship between the daily operating conditions of each sold truck and the corresponding static configuration data. The matching model building module is used to build a configuration condition matching model by matching the daily operating conditions of all sold trucks with the corresponding static configuration data. The configuration recommendation module is used to intelligently recommend corresponding static configuration data for user trucks based on the actual daily operating conditions of the trucks through a configuration working condition matching model.

[0013] Embodiments of the present invention also provide a computer device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the above-described intelligent truck configuration recommendation method.

[0014] Embodiments of the present invention also provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described intelligent truck configuration recommendation method.

[0015] The intelligent truck configuration recommendation method provided by this invention has at least the following beneficial effects: This invention constructs a configuration-condition matching model by mining the matching relationship between truck configuration and operating conditions. This model is built based on the matching relationship between the daily operating conditions of all sold trucks and the corresponding static configuration data. It can intelligently recommend matching vehicle configurations (i.e., static configuration data of trucks) for potential customers. Compared with configuration recommendations based on the personal experience of sales personnel, both efficiency and accuracy can be greatly improved.

[0016] In mining the matching relationship between static configuration data and daily operating conditions of trucks, the daily operating conditions and corresponding static configuration data of several sold trucks are obtained. Considering that truck configuration data (such as gearbox model and rear axle ratio) are discrete, high-dimensional categorical variables, while operating condition data (such as average gradient and altitude) are continuous numerical variables, simply throwing all features into a single model makes it difficult for the model to efficiently learn the complex combination rules of discrete features, and the interpretability is poor. This invention is based on a hierarchical hybrid model architecture. The random forest model, due to its strength in handling high-dimensional discrete features and feature importance evaluation, is used to learn the intrinsic relationship between static configuration combinations (such as the pairing preference between "a certain rear axle ratio" and "a certain gearbox model"). The gradient boosting decision tree model, due to its strong nonlinear fitting ability for continuous numerical features, is used to learn the complex mapping between dynamic operating conditions and vehicle performance indicators (such as fuel consumption and failure rate), thereby obtaining the truck's configuration features and operating condition features respectively.

[0017] Meanwhile, after obtaining the truck's configuration and operating condition characteristics, this invention does not use them directly. Instead, it first determines the rationality evaluation score of different configuration parameter combinations based on the inherent relationship between them. This score is then used as a new enhanced feature and input into the gradient boosting decision tree model along with the original operating condition characteristics. This process obtains the matching relationship between the truck's daily operating conditions and the corresponding static configuration data. Based on this, the configuration and operating condition matching model can intelligently recommend more reasonable truck configurations to users. Attached Figure Description

[0018] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this invention, illustrate exemplary embodiments of the invention and are used to explain the invention, but do not constitute an undue limitation of the invention. In the drawings:

[0019] Figure 1A flowchart illustrating a truck configuration intelligent recommendation method provided by the present invention; Figure 2 A schematic diagram of the overall system architecture and data processing flow provided by the present invention; Figure 3 This is a schematic diagram of the workflow of a configuration condition matching model provided by the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0021] The technical solutions provided by the various embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0022] One embodiment of the present invention relates to a method for intelligent recommendation of truck configurations. The specific process of the intelligent recommendation method for truck configurations in this embodiment can be as follows: Figure 1 As shown, it includes: Step 101: Obtain the daily operating conditions and corresponding static configuration data of several sold trucks.

[0023] Step 102: Extract the intrinsic relationship between different combinations of configuration parameters in the static configuration data of each sold truck using a random forest model, and extract the mapping relationship between the daily operating conditions and truck performance indicators of each sold truck using a gradient boosting decision tree model.

[0024] Step 103: Based on the inherent relationship between different configuration parameter combinations, determine the rationality evaluation score of different configuration parameter combinations, and input the rationality evaluation score of each configuration parameter combination of each sold truck and the mapping relationship between daily operating conditions and truck performance indicators into the gradient boosting decision tree model to obtain the matching relationship between the daily operating conditions of each sold truck and the corresponding static configuration data.

[0025] Step 104: Construct a configuration condition matching model by matching the daily operating conditions of all sold trucks with the corresponding static configuration data.

[0026] Step 105: Based on the actual daily operating conditions of the user's truck, intelligently recommend the corresponding static configuration data of the truck by configuring the working condition matching model.

[0027] The following is a detailed description of the implementation details of the intelligent truck configuration recommendation method in this embodiment. The following content is only for the convenience of understanding and is not necessary for implementing this solution.

[0028] In step 101, regarding static configuration data, the vehicle sales order system and production execution system are first connected. The configuration parameters of the vehicles sold and produced (trucks in this invention), including fuel type, application type, model series, horsepower, version, version, gearbox model, rear axle ratio, tire model, drive type, etc., are standardized and stored in a structured manner to ensure data integrity and uniformity, so as to facilitate subsequent data collection, clustering and retrieval applications.

[0029] The example is explained below: Configuration 1: Fuel type, classification: diesel, CNG, LNG, new energy; Configuration 2: Application type, category: tractor, dump truck, cargo truck, special vehicle; Configuration 3: Vehicle series, categorized as: Series 1, Series 2, Series 3, Series 4, ...; Configuration 4: Horsepower, Classification: 280, 300, 360, 400, 480, 560, 610, 630, ...; Configuration 5: Version type, classification: Lightweight version, Enhanced version, Composite version, Dangerous Goods transport version, ...; Configuration 6: Version, Category: Elite Edition, Classic Edition, Flagship Edition; Configuration 7: Transmission model, category: Model 1, Model 2, Model 3, Model 4, ...; Configuration 8: Rear axle ratio, categories: 2.4, 2.643, 3.545, 3.9, ...; Configuration 9: Tire type, categorized as: Type 1, Type 2, Type 3, Type 4, ...; Configuration 10: Drive type, classification: 4×2, 6×2, 6×4, 8×4, ...

[0030] It should be noted that the above configuration parameters are examples, and configuration parameter maintenance and categorization of each configuration parameter can be maintained as needed.

[0031] Regarding the daily operating conditions of vehicles, the operating conditions are standardized and classified according to road grade, altitude and terrain, slope, load status, ambient temperature, and average speed, so as to combine with configuration data for further data collection, clustering, and retrieval and application.

[0032] The example is explained below: Operating Condition 1: Road Grade, Classification: Expressway, National Highway, Provincial Highway, County Road, General Road, ...; Operating Condition 2: Altitude and Topography, Classification: Plains, Mountains, Hills, Plateaus, ...; Working condition 3: Slope, classified as: flat slope, gentle slope, sloping slope, steep slope, ...; Operating Condition 4: Load Condition, Classification: Heavy Load, Standard Load, Half Load, Unloaded, ...; Operating Condition 5: Ambient temperature, classified as: cold, low temperature, mild, high temperature, hot, ...; Operating Condition 6: Average vehicle speed, classified as: <30km / h, 30~50km / h, 50~70km / h, >70km / h, ...

[0033] It should be noted that the above operating condition items are examples, and maintenance of operating condition items and maintenance of each operating condition category are supported as needed.

[0034] Then, based on the previously structured storage of vehicle static configuration data, combined with vehicle network dynamic operation data, the daily operating conditions of the vehicles are analyzed, identified, and labeled, covering both static configuration and dynamic operating conditions, thereby obtaining the daily operating conditions and corresponding static configuration data of several sold trucks.

[0035] The example is explained below: Tag 1: Configuration tag 1 + operating condition tag 1; Tag 2: Configuration tag 2 + operating condition tag 2; Tag 3: Configuration tag 3 + operating condition tag 3; Tag 4: Configuration tag 4 + operating condition tag 4; Tag 5: Configuration tag 5 + operating condition tag 5; ... Tag n: Configuration tag n + operating condition tag n.

[0036] The configuration label is obtained based on the specific configuration information of the vehicle's sales and production, while the operating condition label is dynamically analyzed and identified based on the vehicle's connected vehicle operation data. The schematic logic is as follows: Road classification: Based on the BeiDou positioning data collected by vehicle networking devices and the road network map data, analysis, comparison and labeling are performed.

[0037] Altitude and terrain: Based on the elevation data of Beidou positioning collected by the vehicle networking equipment, the altitude and terrain where the vehicle travels are classified and labeled according to their proportions: plains: below 200 meters, hills: 200-500 meters, mountains: 500-2000 meters, plateaus: above 2000 meters.

[0038] Slope: Based on the elevation changes collected by the vehicle networking equipment from the Beidou positioning, the slope of the vehicle's driving is classified and labeled according to the proportion. Slope classification: I. 0-5° is flat slope; II. 6-15° is gentle slope; III. 16-25° is sloping slope; IV. 26-35° is steep slope; V. 36-45° is sharp slope; VI. Above 46° is dangerous slope.

[0039] Load status: The vehicle load status is calculated based on data such as fuel consumption and torque collected by the vehicle networking equipment, and then classified and labeled according to the proportion: heavy load, standard load, half load, and no load.

[0040] Operating ambient temperature: Based on the vehicle operating ambient temperature collected by the vehicle environmental monitoring equipment and vehicle networking equipment, the temperature is classified and labeled according to the proportion: cold (<0℃), low temperature (0~10℃), mild (10~25℃), high temperature (25~35℃), hot (>35℃).

[0041] Average vehicle speed: Based on the vehicle speed collected by the vehicle network equipment, the average speed of the vehicle is classified and labeled according to the proportion: <30km / h, 30~50km / h, 50~70km / h, >70km / h.

[0042] In steps 102 and 103, this embodiment implements a hybrid model design based on the heterogeneity of multi-source features. Truck configuration data (such as gearbox model and rear axle ratio) are discrete, high-dimensional categorical variables, while operating condition data (such as average gradient and altitude) are continuous numerical variables. The conventional approach is to simply feed all features into a single model, such as a single Gradient Boosting Decision Tree (GBDT) model. However, this model struggles to efficiently learn complex combination rules in discrete features and suffers from poor interpretability. This embodiment addresses this by employing a hierarchical hybrid model architecture: the Random Forest model, due to its strength in handling high-dimensional discrete features and feature importance assessment, is used to learn the intrinsic relationships between static configuration combinations (e.g., the pairing preference between "a certain rear axle ratio" and "a certain gearbox model"). The Gradient Boosting Decision Tree model, due to its strong nonlinear fitting ability for continuous numerical features, is used to learn the complex mapping between dynamic operating conditions and vehicle performance indicators (such as fuel consumption and failure rate). Furthermore, the fusion stage does not simply average the predictions of the two models. Instead, the evaluation score on "configuration combination rationality" output by the random forest is used as a new strong feature, which is then input into the GBDT model along with the original working condition features for final comprehensive performance prediction. This architecture respects the heterogeneity of the data source, allowing the most suitable model to handle the most suitable data, achieving a 1+1>2 effect.

[0043] In steps 104 and 105, a configuration condition matching model can be constructed by matching the daily operating conditions of all the sold trucks with the corresponding static configuration data. This allows for intelligent truck configuration recommendations.

[0044] In some embodiments, based on the truck's R&D design data, the matching relationship between the daily operating conditions of all sold trucks and the corresponding static configuration data is optimized; a configuration condition matching model is constructed through the optimized matching relationship between the daily operating conditions of all sold trucks and the corresponding static configuration data. The R&D design data includes the engineering boundary design for the normal operation of the truck, including parameters such as integrated engine torque curves (e.g., peak torque 1600 N·m @ 1200 rpm), chassis ultimate load (e.g., drive axle design load 16T), and thermal management thresholds (e.g., radiator operating range -40℃ to 125℃), along with synchronous laboratory simulation reports (e.g., turbocharger failure probability model under high-altitude conditions).

[0045] Specifically, conventional models, with the sole objective of minimizing prediction errors (such as fuel consumption prediction errors), may recommend configurations that theoretically have low fuel consumption but violate common sense in engineering (such as using low horsepower to pull a heavy load up a long slope). This embodiment provides a knowledge-guided loss function: during model construction, we introduce physical constraints based on R&D design data into the loss function. For example, when the model's recommended "horsepower-gradient-load" combination exceeds the chassis system's design limits, a large penalty term is applied to the loss function, thereby guiding the model to learn within feasible engineering boundaries and avoiding recommending unsafe configurations.

[0046] In some embodiments, when making intelligent recommendations for truck configurations through a configuration working condition matching model, the configuration working condition matching model first intelligently recommends static configuration data of multiple trucks based on the actual daily operating conditions of the user's trucks. Then, based on the user's truck purchase needs, the weights of the three truck purchase objectives—economy, reliability, and purchase cost—are adjusted. Based on the three truck purchase objectives with adjusted weights, static configuration data that meets the truck purchase needs is selected from the multiple intelligently recommended static configuration data.

[0047] Specifically, in this embodiment, the model's final output is not a single "optimal solution," but rather a multi-objective optimization ranking of multiple candidate configurations (i.e., Pareto front ranking). A comprehensive scoring function is defined, aiming to simultaneously optimize economy, reliability (estimated failure rate), and acquisition cost. Users can adjust the weights of these three objectives, and the model will recalculate in real time and provide recommendations that best match their current preferences (e.g., a top 3 recommendation). This process transforms the fuzzy concept of "optimum" into an interpretable and customizable multi-objective decision.

[0048] In some embodiments, after extracting the mapping relationship between the daily operating conditions and truck performance indicators of each sold truck through a gradient boosting decision tree model, the performance indicators of the truck under each daily operating condition are predicted based on the mapping relationship. Then, based on the degree of deviation between the predicted performance indicators of the truck under each daily operating condition and the actual performance indicators of the truck under the corresponding daily operating condition, the weight of the matching relationship between the daily operating conditions of each sold truck and the corresponding static configuration data is determined, so as to construct a configuration condition matching model through the matching relationship between the daily operating conditions of sold trucks and the corresponding static configuration data with different weights.

[0049] Specifically, traditional models are fixed once trained and cannot be updated based on newly generated vehicle network data, making them prone to becoming outdated. This embodiment establishes a closed-loop system for continuous learning. Each recommended configuration given to a customer and its subsequent actual operating data (such as real fuel consumption and fault records) are considered as a "state-action-reward" sequence. The system periodically (e.g., weekly) analyzes the deviation between the actual performance of the recommended configuration and the model's predicted values. When vehicles in a certain operating condition dimension (e.g., "altitude") generally exhibit performance indicators (e.g., fuel consumption) that are biased, a reinforcement learning process is automatically triggered. When the deviation of the corresponding operating condition feature within the model from the predicted value exceeds 15%, the system will weight (e.g., significantly increase the importance of the "altitude" feature in the model's decision-making) and fine-tune the model. This allows the model to adapt to market changes and vehicle technology iterations, becoming a "living" and constantly evolving system, which is fundamentally different from a static, one-off model. A quantitative relationship matrix between configuration and operating condition is established with the vehicle's total cost of ownership (TCO) and failure rate as optimization objectives. Example: In high-altitude mountainous areas and under heavy-load conditions, vehicles with a rear axle ratio >3.7 can achieve an average fuel consumption reduction of approximately 8% compared to vehicles with a rear axle ratio of 2.8 (based on vehicle network data statistics). However, a high-horsepower engine (480Ps or more) is required to avoid insufficient power. When the vehicle network reports that the actual fuel consumption deviates from the predicted value by >15% (e.g., actual fuel consumption at high altitude is 38L vs. predicted 32L), dynamic adjustment of feature weights is automatically triggered (increasing the "altitude" feature weight to 70%).

[0050] In one example, an incremental training pipeline (Kafka + Flink stream processing) is deployed to receive and process newly added operational data from the vehicle network in real time (such as fuel consumption and fault codes of newly delivered vehicles), triggering incremental model training. When the actual operational indicators recommended by the model (such as fuel consumption per 100 kilometers) deviate from the predicted value by more than 15%, abnormal cases are automatically marked and model tuning is initiated.

[0051] In one example, an engineering experience rule base is set up (such as "disable supercharged engine when altitude > 3000m"). When the model output violates the rule, the recommended results are automatically frozen and a review work order is pushed to the R&D engineer.

[0052] In some embodiments, based on optimization suggestions from customer feedback regarding successful configuration combination cases in historical sales orders, the matching relationship between the daily operating conditions of all sold trucks and their corresponding static configuration data is optimized. A configuration matching model is then constructed using the optimized matching relationship between the daily operating conditions of all sold trucks and their corresponding static configuration data. This includes successful configuration combination cases in historical sales orders (such as best-selling configurations for heavy-duty tractors in high-altitude areas) and a database of configuration optimization suggestions from customers (such as the requirement for fuel tank heating in cold-region applications). This data provides successful experiences and customer needs from actual business operations, helping the model learn matching rules that better align with market trends.

[0053] In other words, this invention, based on the innovative integration and mining application of multi-source heterogeneous data (pre-marketing business data, R&D design data, and after-market vehicle networking operation data), creates a standardized intelligent vehicle selection tool. This tool is designed to improve truck marketing efficiency and the accuracy of recommended configurations. The aim is to integrate structured static truck configuration data, standardized vehicle operating conditions, and dynamic vehicle networking operation data analysis, combining this information and technological innovation with real-time analysis and judgment of customer vehicle purchase and operation needs to identify the optimal recommended configuration combination. This results in a standardized intelligent vehicle selection and marketing product tool, achieving a dual improvement in truck marketing efficiency and recommendation accuracy. The data processing flow can be found in [reference needed]. Figure 2 .

[0054] First, combining vehicle R&D design data, marketing experience data, and tagged data from vehicle networking data analysis, a configuration condition matching algorithm model is established using the method described in the above embodiments. This model determines the matching relationship between vehicle configuration and suitable operating conditions, and is dynamically adjusted and improved based on feedback from vehicle networking data after application. The workflow of the configuration condition matching model can be found in [reference needed]. Figure 3 .

[0055] Then, through the application of information technology, the above-mentioned configuration condition matching algorithm model is transformed into a digital and standardized vehicle selection tool. It supports the custom selection of the vehicle's intended operating conditions and the setting of different operating condition weights. Combined with user input settings, the configuration condition matching algorithm model is invoked to dynamically calculate and match recommended configurations and display them according to priority. At the same time, combined with vehicle network data, the operational performance of vehicles with the same configuration that have been sold is filtered and displayed to help customers understand and promote ordering, thereby improving marketing efficiency and the accuracy of recommended configurations.

[0056] This car selection tool includes: (1) Intelligent Interactive System: ① Operating Condition Configuration Device: It supports multi-level working condition definition (basic layer: road grade + load status; advanced layer: superimposed temperature zone + humidity range), and the weight slider can be freely adjusted (such as "slope weight 60% + temperature weight 20% + altitude weight 20%)".

[0057] Built-in industry scenario template library (one-click loading of preset schemes for "Coal Transportation: Heavy Load + Mountain Road + Dust Environment").

[0058] ② Real-time decision engine: The matching model from step four is invoked, and the top 3 recommended configurations are returned within 200ms after the user adjusts the parameters, sorted by comprehensive score (economic weight 40% + reliability weight 35% + cost weight 25%).

[0059] (2) Data-driven recommendation dashboards: ① Multi-dimensional comparison view: The key parameters of the recommended configurations are compared horizontally (such as engine horsepower, rear axle ratio, and fuel tank capacity), and the differentiated advantages are marked (such as "Configuration A has 5% lower fuel consumption but costs 8,000 more").

[0060] Embedded operational data dashboard with the same configuration: Based on real-time vehicle network data, it displays the operational indicators of vehicles with the same configuration that have been sold (such as actual average fuel consumption, average monthly attendance rate, and distribution of typical faults).

[0061] Example: When a user selects "tractor + standard load + high altitude", the system displays a fuel consumption curve of 50 vehicles with the same configuration on the Qinghai-Tibet Highway (median 32L / 100km).

[0062] (3) Decision support and risk verification: ①Simulated Operations Calculator: Input user waybill parameters (annual mileage, fuel price, toll fees, etc.) to automatically generate a 5-year TCO comparison report with different configurations.

[0063] ②Configure compatibility verification: Real-time detection of the feasibility of user-defined configuration combinations (such as "low-horsepower engine + high-ratio rear axle" may trigger a power shortage alarm) and recommendation of alternative solutions.

[0064] ③ Interpretability module: The application of SHAP value analysis reveals key influencing factors (e.g., in high-altitude scenarios, the "rear axle ratio" contributes 65% to fuel consumption).

[0065] (4) System integration and expansion: ① API integration capability: The configuration recommendation interface is open to the dealer's DMS order system and the official website's vehicle selection page, and supports calls from third-party systems.

[0066] ②Data Cockpit: Real-time monitoring of the adoption rate of recommended solutions nationwide (e.g., the adoption rate of tractor solutions in North China is 92.3%), customer satisfaction (NPS≥85), and operational indicator deviation warnings (fuel consumption deviation >10% is automatically highlighted in red).

[0067] This embodiment can match and recommend configurations for the customer's intended operating conditions, improving marketing efficiency and recommendation accuracy. Through data fusion innovation and algorithm model breakthroughs, it solves the fundamental contradiction in configuration recommendations in the commercial vehicle industry—the mismatch between complex operating conditions and limited human experience. The technological value lies in the paradigm shift from "experience-driven" to "data intelligence-driven".

[0068] The intelligent truck configuration recommendation method of the present invention has the following beneficial effects: 1) Significantly improved accuracy and reliability of recommendations: In internal testing, the accuracy of configuration recommendations for complex operating conditions (such as high-altitude heavy-load and cold-region transportation) has increased from about 50% driven by experience to over 92%, and can effectively avoid configuration combinations that violate engineering principles, significantly reducing the early failure rate of vehicles.

[0069] 2) Marketing decision-making efficiency has achieved an order of magnitude leap: The average time for selecting vehicle configuration has been shortened from more than 3 hours to less than 10 minutes, which greatly improves marketing efficiency and customer experience and solves the professional bottleneck of novice sales consultants.

[0070] 3) Possesses continuous self-optimization capabilities and responds agilely to market changes: The model can continuously optimize based on feedback data from the Internet of Vehicles, and the iteration cycle of the recommendation strategy is shortened from the traditional quarterly / annual cycle to the weekly cycle, demonstrating agility in responding to new transportation routes and new vehicle model technology iterations.

[0071] 4) Achieve digital accumulation and standardization of core enterprise knowledge: Solidify valuable expert experience into the algorithm rule base, reduce the enterprise's reliance on individual experience, realize the digital inheritance and standardized application of car selection knowledge, and build a sustainable differentiated competitiveness.

[0072] The steps of the various methods described above are only for clarity. In practice, they can be combined into one step or some steps can be split into multiple steps. As long as they include the same logical relationship, they are all within the protection scope of this invention. Adding insignificant modifications or introducing insignificant designs to the algorithm or process, without changing the core design of the algorithm and process, are also within the protection scope of this invention.

[0073] Another embodiment of the present invention relates to a truck configuration intelligent recommendation system. The implementation details of this truck configuration intelligent recommendation system are described below. The following implementation details are provided for ease of understanding and are not essential for implementing this solution. The truck configuration intelligent recommendation system of this embodiment includes: The data acquisition module is used to acquire the daily operating conditions and corresponding static configuration data of several sold trucks; The feature extraction module is used to extract the inherent relationship between different combinations of configuration parameters in the static configuration data of each sold truck through a random forest model, and to extract the mapping relationship between the daily operating conditions and truck performance indicators of each sold truck through a gradient boosting decision tree model. The matching relationship acquisition module is used to determine the reasonable evaluation score of different configuration parameter combinations based on the inherent relationship between different configuration parameter combinations, and input the evaluation score of each configuration parameter combination of each sold truck and the mapping relationship between daily operating conditions and truck performance indicators into the gradient boosting decision tree model to obtain the matching relationship between the daily operating conditions of each sold truck and the corresponding static configuration data. The matching model building module is used to build a configuration condition matching model by matching the daily operating conditions of all sold trucks with the corresponding static configuration data. The configuration recommendation module is used to intelligently recommend corresponding static configuration data for user trucks based on the actual daily operating conditions of the trucks through a configuration working condition matching model.

[0074] It is not difficult to see that this embodiment is a system embodiment corresponding to the above method embodiments, and this embodiment can be implemented in conjunction with the above method embodiments. The relevant technical details and technical effects mentioned in the above embodiments are still valid in this embodiment, and will not be repeated here to reduce repetition. Accordingly, the relevant technical details mentioned in this embodiment can also be applied to the above embodiments.

[0075] It is worth mentioning that all modules involved in this embodiment are logical modules. In practical applications, a logical unit can be a physical unit, a part of a physical unit, or a combination of multiple physical units. Furthermore, to highlight the innovative aspects of this invention, this embodiment does not introduce units that are not closely related to solving the technical problem proposed by this invention; however, this does not mean that other units are absent from this embodiment.

[0076] Another embodiment of the present invention relates to a computer device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the intelligent truck configuration recommendation method of the above embodiments.

[0077] The memory and processor are connected via a bus, which can include any number of interconnecting buses and bridges, connecting various circuits of one or more processors and memories. The bus can also connect various other circuits, such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and will not be described further herein. The bus interface provides an interface between the bus and the transceiver. The transceiver can be a single element or multiple elements, such as multiple receivers and transmitters, providing a unit for communicating with various other devices over a transmission medium. Data processed by the processor is transmitted over the wireless medium via an antenna, which further receives data and transmits it to the processor.

[0078] The processor manages the bus and general processing, and also provides various functions, including timing, peripheral interfaces, voltage regulation, power management, and other control functions. Memory is used to store data used by the processor during operation.

[0079] Another embodiment of the present invention relates to a computer-readable storage medium storing a computer program. When executed by a processor, the computer program implements the method embodiments described above.

[0080] That is, those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware. This program is stored in a storage medium and includes several instructions to cause a device (which may be a microcontroller, chip, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0081] Those skilled in the art will understand that the above embodiments are specific embodiments for implementing the present invention, and in practical applications, various changes can be made to them in form and detail without departing from the spirit and scope of the present invention.

Claims

1. A method for intelligent recommendation of truck configuration, characterized in that, The method includes: Obtain daily operating conditions and corresponding static configuration data for several sold trucks; The inherent relationship between different combinations of configuration parameters in the static configuration data of each sold truck is extracted by random forest model, and the mapping relationship between the daily operating conditions and truck performance indicators of each sold truck is extracted by gradient boosting decision tree model. Based on the inherent relationship between different configuration parameter combinations, the rationality evaluation score of different configuration parameter combinations is determined. The rationality evaluation score of each configuration parameter combination of each sold truck and the mapping relationship between daily operating conditions and truck performance indicators are input into the gradient boosting decision tree model to obtain the matching relationship between the daily operating conditions of each sold truck and the corresponding static configuration data. A configuration condition matching model is constructed by matching the daily operating conditions of all sold trucks with the corresponding static configuration data. By configuring a working condition matching model, intelligent recommendations are made for the corresponding static configuration data of the user's truck based on the actual daily operating conditions of the truck.

2. The intelligent truck configuration recommendation method according to claim 1, characterized in that, The process involves constructing a configuration condition matching model by matching the daily operating conditions of all sold trucks with their corresponding static configuration data, including: Based on the mapping relationship between the daily operating conditions and truck performance indicators of each sold truck extracted by the gradient boosting decision tree model, the performance indicators of the truck under each daily operating condition are predicted. Based on the degree of deviation between the predicted performance indicators of the truck under each daily operating condition and the actual performance indicators of the truck under the corresponding daily operating condition, the weight of the matching relationship between the daily operating conditions of each sold truck and the corresponding static configuration data is determined. A configuration condition matching model is constructed by matching the daily operating conditions of sold trucks with different weights with the corresponding static configuration data.

3. The intelligent truck configuration recommendation method according to claim 1, characterized in that, The process of training and constructing a configuration condition matching model by matching the daily operating conditions of all sold trucks with the corresponding static configuration data includes: Based on the truck's R&D and design data, the matching relationship between the daily operating conditions of all sold trucks and the corresponding static configuration data is optimized; among which, the R&D and design data includes the engineering boundary design to ensure the normal operation of the truck. A configuration condition matching model is constructed by matching the optimized daily operating conditions of all sold trucks with the corresponding static configuration data.

4. The intelligent truck configuration recommendation method according to claim 1, characterized in that, The process involves constructing a configuration condition matching model by matching the daily operating conditions of all sold trucks with their corresponding static configuration data, including: Based on customer feedback on configuration combination optimization suggestions from successfully verified configuration combination cases in historical sales orders, the matching relationship between the daily operating conditions and corresponding static configuration data of all sold trucks is optimized. A configuration condition matching model is constructed by matching the optimized daily operating conditions of all sold trucks with the corresponding static configuration data.

5. The intelligent truck configuration recommendation method according to claim 1, characterized in that, The intelligent recommendation of corresponding truck static configuration data based on the actual daily operating conditions of the user's truck by configuring a working condition matching model includes: By configuring a working condition matching model, the system intelligently recommends static configuration data for multiple trucks based on the actual daily operating conditions of the user's trucks. Based on the user's truck purchase needs, the weights of the three truck purchase objectives—economy, reliability, and purchase cost—are adjusted. Then, based on these adjusted weights, static configuration data that meets the truck purchase needs is selected from multiple intelligently recommended static configuration data.

6. The intelligent truck configuration recommendation method according to any one of claims 1 to 5, characterized in that, The static configuration data includes the following configuration parameters: fuel type, application type, vehicle series, horsepower, model, version, transmission model, rear axle ratio, tire model, and drive type.

7. The intelligent truck configuration recommendation method according to any one of claims 1 to 5, characterized in that, The daily operating conditions are determined by the following truck operating data: road grade, altitude, gradient, load condition, ambient temperature, and average speed.

8. A truck configuration intelligent recommendation system, characterized in that, The system includes: The data acquisition module is used to acquire the daily operating conditions and corresponding static configuration data of several sold trucks; The feature extraction module is used to extract the inherent relationship between different combinations of configuration parameters in the static configuration data of each sold truck through a random forest model, and to extract the mapping relationship between the daily operating conditions and truck performance indicators of each sold truck through a gradient boosting decision tree model. The matching relationship acquisition module is used to determine the reasonable evaluation score of different configuration parameter combinations based on the inherent relationship between different configuration parameter combinations, and input the evaluation score of each configuration parameter combination of each sold truck and the mapping relationship between daily operating conditions and truck performance indicators into the gradient boosting decision tree model to obtain the matching relationship between the daily operating conditions of each sold truck and the corresponding static configuration data. The matching model building module is used to build a configuration condition matching model by matching the daily operating conditions of all sold trucks with the corresponding static configuration data. The configuration recommendation module is used to intelligently recommend corresponding static configuration data for user trucks based on the actual daily operating conditions of the trucks through a configuration working condition matching model.

9. A computer device, characterized in that, include: At least one processor; And a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the truck configuration intelligent recommendation method as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the intelligent truck configuration recommendation method as described in any one of claims 1 to 7.

Images