A power control method for turbojet snowplow based on artificial intelligence

Abstract

The present invention discloses an artificial intelligence-based power control method for a turbojet snowplow, which relates to the technical field of engineering equipment and includes the following steps: S1, performing scene type image recognition; S2, mapping an operating environment sensitivity vector to quantify the sensitivity of environmental factors; S3, constructing a primary power strategy function to generate a preliminary jet power strategy; S4, adaptively adjusting the power and outputting the adjusted control strategy; S5, performing time-domain segmented control of the turbojet engine power output; and S6, interrupting and reverting the segmented-controlled turbojet engine power in abnormal situations. This artificial intelligence-based power control method for a turbojet snowplow, by integrating scene image semantic recognition, environmental sensitivity modeling, power strategy function construction, and an adaptive dynamic adjustment mechanism, not only overcomes the technical bottleneck of the inflexible power control of existing equipment but also achieves precise matching of power output with the actual operating environment.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a power control method for a turbojet snow remover based on artificial intelligence. Background Art

[0002] With the rapid development of infrastructure, the massive construction of highways, elevated bridges, and various airports has brought higher requirements for winter snow removal efficiency and operational safety. Traditional snow removal methods mainly include manual snow shoveling, mechanical snow pushing, and chemical snow melting. However, manual methods are inefficient, the mechanical equipment is bulky, and the snow removal is not thorough. Chemical methods, while effective quickly, are prone to environmental pollution and road corrosion. In addition, existing turbojet snow removal equipment mostly uses retired aircraft engines, which have defects such as difficult maintenance, high fuel consumption, and unadjustable power, making it difficult to adapt to changing operating scenarios and complex environmental conditions. In actual use, these devices are often unable to implement dynamic adjustments based on differences in snow distribution, wind speed changes, or different road materials, resulting in energy waste and poor snow removal results. Especially in complex scenarios, existing control systems lack intelligent control capabilities based on comprehensive perception of image semantics and environmental factors, and are unable to achieve refined and adaptive adjustment of jet power. Summary of the Invention

[0003] In view of the shortcomings of the existing technology, the present invention provides a power control method of a turbojet snow blower based on artificial intelligence to solve the problems raised in the above background technology.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] In a first aspect, an embodiment of the present invention provides a power control method for a turbojet snow blower based on artificial intelligence, comprising the following steps:

[0006] S1. Perform scene type image recognition, use feature vectors to map and compare with semantic label sets, and output the semantic region level identification of the current scene;

[0007] S2. Map the output semantic region level identifier to its corresponding working environment sensitivity vector to quantify the sensitivity of the region to environmental factors such as temperature and wind speed;

[0008] S3. Based on the output working environment sensitivity vector, construct a primary power strategy function and generate a preliminary jet power strategy;

[0009] S4. Based on the preliminary jet power strategy, the power is adaptively adjusted by detecting the residual snow thickness and surface temperature rise in the snow removal area, and the adjusted control strategy is output;

[0010] S5. Based on the output control strategy, the turbojet engine power output is controlled in a time domain segmented manner;

[0011] S6. Perform interruption and fallback operations on the segmented-controlled turbojet engine power in abnormal situations.

[0012] To further optimize this technical solution, in step S1, real-time image data of the operation scene is collected by the front-end camera module and the geo-tagging system of the turbojet snow blower, and feature vectors of the operation area are extracted based on the multi-level convolutional attention network of artificial intelligence, including surface texture, structural boundaries, and facility types;

[0013] Construct an image scene recognition function, use the feature vector to map and compare with the semantic label set, and output the semantic area level identification value of the current scene.

[0014] To further optimize this technical solution, the image scene recognition function is as follows:

[0015] ;

[0016] in,

[0017] : The semantic area level identifier of the current scene, including airport runways, viaducts, and suburban roads;

[0018] : original image input;

[0019] : The semantic features extracted from the image by the i-th feature channel;

[0020] : The importance weight of each semantic feature;

[0021] : Post-processing mapping function, used to classify the multi-feature combination results into discrete semantic levels;

[0022] The scene recognition result of the image scene recognition function provides structured prior knowledge for subsequent power adjustment.

[0023] Further optimizing this technical solution, in step S2, each scene type corresponds to a set of environmental sensitivity features, and the semantic area level identifiers are automatically mapped to the work environment sensitivity vector through the parameter database to construct an environmental response factor model. This model is not only based on static standards, but also introduces current real-time meteorological data to perform weighted correction on the work environment sensitivity vector;

[0024] The environmental response factor model is as follows:

[0025] ;

[0026] in,

[0027] : Operating environment sensitivity vector, used to guide subsequent power control;

[0028] :According to semantic region level identification Preset environment response template vector;

[0029] :Based on current real-time weather data A correction function for correcting the environmental sensitivity vector;

[0030] The model builds a logical association bridge between semantic types and real-time physical environments, which is used for continuous mapping from image recognition to environmental constraint modeling.

[0031] Further optimizing the technical solution, in step S3, after obtaining the operating environment sensitivity vector, a primary power strategy function is generated by setting a logic builder in combination with the response capability of the available jet modules onboard the turbojet snow blower, so that the jet output can effectively remove snow without causing excessive thermal stress, wind erosion, or local damage to the operating area;

[0032] The primary power strategy function is as follows:

[0033] ;

[0034] in,

[0035] : Candidate jet power control value, the unit is not fixed, indicating the relative level;

[0036] : The three main dimensions of the operating environment sensitivity vector include temperature sensitivity, wind speed adaptability, and structural response limit;

[0037] : Primary power control function based on operating environment sensitivity vector;

[0038] : System preset weights are set based on device performance and historical experience;

[0039] The output of this function is Instead of making a final control output, a candidate power strategy that can be optimized is constructed, and adaptive power adjustment is performed in subsequent steps.

[0040] Further optimizing the technical solution, in step S4, the turbojet snow remover is equipped with a surface feedback sensor module, including a residual snow imaging detector, an infrared temperature sensor, and a micro laser ranging device; the residual snow thickness, surface humidity, and temperature rise curve of the operating area are collected once per second to determine whether the current jet power reaches the ideal snow removal effect and the heat impact threshold;

[0041] Feedback mechanism with candidate jet power control value With the detection data as input and the control condition, a snow state adaptation function is constructed to realize power adaptive adjustment and output the actual jet power value to prevent damage to the ground structure caused by insufficient or excessive power.

[0042] To further optimize this technical solution, the snow state adaptation function is as follows:

[0043]

[0044] in,

[0045] : The actual jet power value;

[0046] : The currently detected residual snow thickness;

[0047] : The lower limit of ideal clearing thickness set by the system;

[0048] : The current detected surface temperature;

[0049] : The maximum acceptable surface temperature of the structural material;

[0050] : Regulation factor, representing the intensity of feedback effect.

[0051] Further optimizing the technical solution, in step S5, the turbojet engine power output is controlled in time domain segmentation, the total mission cycle is divided into multiple control segments, and the actual jet power value in each segment is controlled. For nonlinear power allocation, the allocation model is as follows:

[0052]

[0053] in,

[0054] : jet power output of the ith control segment;

[0055] : The total number of segments into which the snow removal operation cycle is divided;

[0056] : Control the increase or decrease of each section based on the energy consumption optimization coefficient extracted from historical mission data;

[0057] The model adopts a nonlinear regulation method based on a sine function to achieve multi-segment adaptive power output driven by task subdivision.

[0058] To further optimize the technical solution, in step S6, the turbojet snow blower receives temperature curves, wind pressure fluctuations, and nozzle abnormal acoustic data in real time, and performs a diagnosis once it is found that the parameters exceed the threshold value;

[0059] According to the diagnosis results, the current power output Make corrections, that is, continue to execute the current power output; or directly jump to the safe fallback power value to avoid irreversible damage such as nozzle erosion and road cracking;

[0060] Safety fallback power value The judgment model is as follows:

[0061]

[0062] in,

[0063] : Back-off coefficient, a constant less than 1, determined based on the feedback from the expert network.

[0064] To further optimize this technical solution, a surface material identification and adaptation mechanism is introduced when performing time-domain segmented control of the turbojet engine power output. Before starting the power segmented control, visual recognition and spectral reflectance analysis are used to determine the type of working pavement material, including concrete, high-strength asphalt, and rubber pavement. Based on the material's thermal expansion coefficient and thermal tolerance limit, the corresponding power curve segmentation template is selected.

[0065] In a second aspect, an embodiment of the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program instructions are executed by the processor, the steps of a power control method of a turbojet snow blower based on artificial intelligence as described in the first aspect of the present invention are implemented.

[0066] In a third aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program instructions are executed by a processor, the steps of a power control method for a turbojet snow blower based on artificial intelligence as described in the first aspect of the present invention are implemented.

[0067] Compared with the prior art, the present invention provides a power control method for a turbojet snow blower based on artificial intelligence, which has the following beneficial effects:

[0068] This AI-based power control method for turbojet snowplows integrates scene image semantic recognition, environmental sensitivity modeling, power strategy function construction, and an adaptive dynamic adjustment mechanism. This method not only overcomes the technical bottleneck of existing equipment's inflexible power control, but also achieves precise matching of power output with the actual operating environment. In complex or variable snowplow scenarios, this method implements time-domain segmented control and abnormal interruption protection for the turbojet engine based on semantic region levels from image recognition and real-time sensor data. This ensures thorough snow removal while effectively reducing energy consumption and equipment wear, improving response efficiency and environmental adaptability. This represents significant technological advancement and practical application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0070] Figure 1 This is a flow chart of a power control method for a turbojet snow blower based on artificial intelligence proposed by the present invention. DETAILED DESCRIPTION

[0071] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0072] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0073] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it constitute a separate or selective embodiment that is mutually exclusive with other embodiments.

[0074] Example 1:

[0075] Reference Figure 1 , which is the first embodiment of the present invention, provides a power control method for a turbojet snow blower based on artificial intelligence, comprising the following steps:

[0076] S1. Perform scene type image recognition, use feature vectors to perform mapping and comparison with semantic label sets, and output the semantic region level identification of the current scene.

[0077] Real-time image data of the operation scene is collected through the front-end camera module and geographic tagging system of the turbojet snow blower, and feature vectors of the operation area, including surface texture, structural boundaries, and facility types, are extracted based on the multi-level convolutional attention network based on artificial intelligence.

[0078] Construct an image scene recognition function, use the feature vector to map and compare with the semantic label set, and output the semantic area level identification value of the current scene.

[0079] Traditional snow removal equipment typically relies on geographic information systems (such as GPS) or pre-set operating paths to categorize the work area. However, its recognition accuracy is limited by spatial resolution and map update frequency, and it cannot perceive on-site changes (such as temporary obstacles or different types of pavement). This step, however, uses AI-powered image recognition to directly identify scene semantics based on real-time images captured by front-end cameras, eliminating the need for external mapping systems or infrastructure positioning modules. This significantly improves adaptability and the real-time nature of scene analysis.

[0080] The image scene recognition function is as follows:

[0081]

[0082] in,

[0083] : The semantic area level identifier of the current scene, including airport runways, viaducts, and suburban roads;

[0084] : Original image input, collected in real time by the front-end camera module;

[0085] : The semantic features extracted from the image by the i-th feature channel, such as edge density, texture coarseness, surface reflectivity, background configuration, etc.

[0086] : The importance weight of each semantic feature is obtained by offline training in the semantic classification task through deep learning methods;

[0087] : Post-processing mapping function, used to classify the multi-feature combination results into discrete semantic levels, all weighted features are processed by the function Classified as a discrete scene semantic level identifier (For example =1 means "city road", =3 for "airport runway");

[0088] The scene recognition results of this image scene recognition function provide structured prior knowledge for subsequent power regulation. Compared to traditional single-image classification methods, this method provides structured prior knowledge for subsequent power regulation. This semantically driven control input does not rely on low-level geometric contour recognition, improving the intelligence of the power regulation response.

[0089] In this embodiment, image acquisition not only relies on a wide-angle camera but also dynamically switches to infrared imaging mode based on ambient illumination conditions, adapting to low-visibility scenarios such as nighttime or snowstorms. Furthermore, the image processing chain is handled in real time by a low-latency edge computing unit, avoiding the latency associated with cloud-based recognition and ensuring sub-second power control response.

[0090] S2. Map the output semantic region level identifier to its corresponding working environment sensitivity vector to quantify the sensitivity of the region to environmental factors such as temperature and wind speed.

[0091] Traditional snow removal equipment control systems often only make rough power configurations based on the geographical location of the operating area or preset task parameters, and lack the ability to respond to the actual state of the environment. This is especially true when encountering special structural areas such as viaducts, tunnel entrances, and slopes. They are prone to ignoring their sensitivity to physical factors such as temperature and wind speed, leading to power configuration imbalances, reduced efficiency, and even road damage. This step, by introducing the mapping logic between semantic area level identifiers and environmental sensitivity vectors and integrating the dynamic correction response characteristics of real-time meteorological data, achieves intelligent coupling between semantic scenarios and environmental physical constraints. It can meticulously depict the tolerance and safety boundaries of different types of areas for turbojet snow removal power, providing more realistic and detailed environmental feedback conditions for subsequent control strategies. This approach significantly improves the scenario adaptability and environmental response accuracy of the control strategy, overcoming the shortcomings of the traditional model of "discrete scenario information and rigid environmental adaptation."

[0092] Each scenario type corresponds to a set of environmental sensitivity features (such as antifreeze requirements, structural deformation sensitivity, and thermal radiation thresholds). The parameter database automatically maps semantic region level identifiers to operational environment sensitivity vectors. This model then constructs an environmental response factor model. This model uses not only static standards but also real-time meteorological data (such as ambient temperature, wind speed, and humidity) to weight and modify the operational environment sensitivity vectors.

[0093] The environmental response factor model is as follows:

[0094]

[0095] in,

[0096] :The operating environment sensitivity vector is used to guide the subsequent power control and reflect the constraints and influences of environmental factors on power control in the operating environment. The operating environment sensitivity vector consists of two parts: one is based on The mapped basic template vector , and secondly, combining real-time meteorological data Calculated , the sum of the two forms the final input vector for power regulation modeling The vector It is specifically used to quantify the response degree to physical factors such as wind speed, temperature and humidity in the current scene, and is the key input basis in step S3.

[0097] :According to semantic region level identification Preset environmental response template vectors (e.g., the default thermal expansion and contraction effect of an elevated bridge is high);

[0098] :Based on current real-time weather data (wind speed, temperature, etc.) to correct the environmental sensitivity vector to ensure that the obtained vector Can reflect the current specific natural environment characteristics.

[0099] First, call the preset scene-environment template dictionary to input the semantic region level identifier Mapped into an environmental response template vector; then collect the meteorological parameters of the current operation area, calculate its impact on various environmental characteristics, and This is added to the original template to obtain the final operational environment sensitivity vector. This vector will be used in subsequent steps to adjust the safe upper limit of turbojet engine power and the power increase rate to ensure that the strategy is physically reasonable and environmentally adaptable.

[0100] The model builds a logical association bridge between semantic types and real-time physical environments, which is used for continuous mapping from image recognition to environmental constraint modeling.

[0101] In practical applications, turbojet snow removal operations often face extremely variable operating environments. For example, elevated bridges are more susceptible to ice formation than the ground in winter. Airport runways are prone to localized snow accumulation due to their open wind fields. Even some tunnel exits may experience the formation of invisible thin ice due to abnormal heat exchange. By establishing a direct mapping between "structural type" and "physical properties," a detailed characterization of environmental sensitivity is achieved. For example, the system automatically assigns parameter vectors such as high wind sensitivity, high cold shrinkage stress response, and high de-icing redundancy requirements to the semantic level of "elevated bridge." Upon detecting real-time changes such as increased wind speed and decreased temperature, the system weights these sensitivity dimensions and dynamically adjusts the power strategy. This deep integration of the physical scenario and snow removal operational requirements enables snow removal equipment to "understand the environment," breaking away from the static and inflexible limitations of traditional power strategies. It achieves a high degree of coupling between scenario-driven and physical constraints, laying the foundation for both data logic and decision-making accuracy in intelligent snow removal systems.

[0102] In actual application, if the current operation area is identified as a "viaduct" type scene, the semantic area level identifier Z is used to represent the viaduct. When executing step S2, the environment response template vector of the scene is extracted from the scene-environment template dictionary:

[0103] e1 (wind speed sensitivity factor): high (e.g. set to 0.8)

[0104] e2 (temperature sensitivity factor): High (e.g. set to 0.75)

[0105] e3 (humidity sensitivity factor): medium (e.g. set to 0.45)

[0106] at this time, The output is an initial vector: [0.8, 0.75, 0.45], which means that the viaduct is extremely sensitive to wind speed and temperature fluctuations, and has a moderate response to humidity.

[0107] Based on the monitoring results of meteorological data, the following data are obtained:

[0108] Wind speeds continued to rise, reaching over 8 meters per second;

[0109] The temperature dropped to -12 degrees Celsius;

[0110] Keep the humidity around 55%.

[0111] At this point, the system calls the correction function , dynamically adjust the original environment response template vector:

[0112] For the wind speed sensitivity factor e1, Increase it from 0.8 to 0.9;

[0113] For the temperature sensitivity factor e2, Increase it from 0.75 to 0.88;

[0114] For the humidity sensitivity factor e3, Maintain the original value or slightly increase it to 0.48.

[0115] The final output environmental response factor vector =[0.9, 0.88, 0.48], this vector is used to adjust specific parameters including power growth gradient, output peak upper limit, jet rate buffer, etc.

[0116] S3. Based on the output working environment sensitivity vector, a primary power strategy function is constructed to generate a preliminary jet power strategy.

[0117] The power control of traditional snow removal equipment mostly relies on static task plans and manual experience adjustments, and is unable to match the complex and changing operating environment characteristics in real time. Especially under the interaction of different regions and weather conditions, it is easy to have "insufficient power output" or "excessive snow removal", which in turn affects operating efficiency and even causes road damage.

[0118] After obtaining the operational environment sensitivity vector, combined with the response capabilities of the available jet modules onboard the turbojet snow blower, a primary power strategy function is generated by setting a logic builder, so that the jet output can effectively remove snow without causing excessive thermal stress, wind erosion, or local damage to the operational area.

[0119] The primary power strategy function is as follows:

[0120]

[0121] in,

[0122] : Candidate jet power control value, the unit is not fixed, indicating the relative level;

[0123] : The three main dimensions of the operating environment sensitivity vector include temperature sensitivity, wind speed adaptability, and structural response limit;

[0124] : Primary power control function based on operating environment sensitivity vector;

[0125] : The system presets weights, which are set based on equipment performance and historical experience, and are adjusted according to different snow removal scenarios and equipment response curves to form a weighted combination;

[0126] The output of this function is Instead of making the final control output, we construct a candidate power strategy that can be optimized and perform adaptive power adjustment in subsequent steps. First, we call the output vector of step S2 , extracting its main dimensions as modeling input; the system selects the appropriate weight coefficient combination based on the equipment type (such as whether it is a dual-nozzle structure), historical feedback performance, and current job priority; and then synthesizes the weighted model value.

[0127] Compared with traditional models, this function not only takes into account meteorological factors such as temperature and wind speed, but also integrates structural sensitivity and regional semantic levels, making the power output scenario-relevant, avoiding the "one-size-fits-all" strategy error, and realizing the intelligent linkage between "environmental understanding-power allocation".

[0128] This function does not directly issue control instructions to the snow removal equipment, but first constructs a power candidate strategy derived from the current environment and area type. Its purpose is to provide a power configuration scheme framework with basic adaptability for the turbojet snow removal system so that more detailed and dynamic adjustments can be made in subsequent steps. For example, when the environmental response factor vector shows that the wind speed in the current area changes drastically and the structure is vulnerable, the system will increase the power of the system. The system will adjust the weight of the primary jet to reduce the primary jet power and prevent the high-energy jet from causing additional damage. On the contrary, in the wide and open airport runway area, if the temperature is low and the humidity is high, and the wind speed is low and the stability is strong, the system will appropriately increase the upper limit of the jet power to enhance the snow removal efficiency.

[0129] S4. Based on the preliminary jet power strategy, the power is adaptively adjusted by detecting the residual snow thickness and surface temperature rise in the snow removal area, and the adjusted control strategy is output.

[0130] In traditional turbojet snow removal systems, once jet power is set, it often remains constant throughout the entire snow removal operation, or relies on manual adjustments based on experience. This approach lacks real-time feedback and can easily lead to insufficient or excessive snow removal, compromising snow removal efficiency and potentially causing thermal damage to the road surface.

[0131] The turbojet snowplow is equipped with a surface feedback sensor module, including a residual snow imaging detector, an infrared temperature sensor, and a micro laser ranging device. It collects data on the residual snow thickness, surface humidity, and temperature rise curve in the operating area once per second to determine whether the current jet power reaches the ideal snow removal effect and the thermal impact threshold.

[0132] Feedback mechanism with candidate jet power control value With the detection data as input and the control condition, a snow state adaptation function is constructed to realize power adaptive adjustment and output the actual jet power value to prevent damage to the ground structure caused by insufficient or excessive power.

[0133] The snow state adaptation function is as follows:

[0134]

[0135] in,

[0136] : The actual jet power value;

[0137] : The currently detected residual snow thickness;

[0138] : The lower limit of ideal clearing thickness set by the system;

[0139] : The current detected surface temperature;

[0140] : The maximum acceptable surface temperature of the structural material;

[0141] : Adjustment factor, representing the feedback strength, learned based on empirical data from historical snow removal tasks.

[0142] The specific usage is: During the process, the thickness of residual snow and the surface temperature of the snow removal area are collected in real time and substituted into the above model to calculate the correction term. > ), the correction term is positive, If the value is raised, the jet power will be enhanced; if the surface temperature is too high ( > ), the correction term is negative, The final result is a set of feedback driven , to achieve dynamic adaptation to different snow removal conditions.

[0143] In practice, high-frequency sampling is performed using surface feedback sensor modules deployed along the turbojet snowplow's operating path. These sensors work together to provide continuous, fine-grained, dynamic monitoring of surface conditions. The system reads sensor data once per second or more, extracting trends in the current residual snow thickness and surface temperature. These data are then compared with preset target values to accurately determine whether jet power needs to be increased or decreased.

[0144] This process doesn't simply rely on a single parameter to make decisions. Instead, the model coordinates the residual snow and thermal impact dimensions, achieving higher sensitivity and more coordinated responses. For example, when the residual snow hasn't completely cleared but the surface temperature approaches a critical value, the system can use the model to maintain the jet power within a safe minimum range, thereby avoiding the risk of thermal damage. Similarly, when the ground heats up slowly under cold wind conditions, the model can also allow for brief power overshoots to ensure operational efficiency.

[0145] S5. Based on the output control strategy, the turbojet engine power output is controlled in time domain segmented manner.

[0146] Traditional turbojet snow removal equipment mostly uses fixed or simple linear segmented power control strategies. Even within these segmented strategies, each segment's output is often preset to a constant value or linearly increasing value, lacking effective response to the dynamic characteristics of mission phases and snow removal scenarios. This approach can easily lead to uneven energy consumption, severe thermal shock, localized overload, and low efficiency. This approach is particularly unsuitable for critical areas such as airport runways and railway lines, where precise timing is crucial.

[0147] The turbojet engine power output is controlled in time domain segmented, and the total mission cycle is divided into multiple control segments. In each segment, the actual jet power value is set. Perform nonlinear power allocation.

[0148] Snow removal tasks are often staged, such as "initial cleaning", "centralized blowing", "residual snow drying" and other different stages. In order to avoid energy consumption and material thermal fatigue problems caused by power fluctuations, this system divides each snow removal cycle into multiple sub-segments. Adjustments are made to make the total output power curve smoother, more energy-efficient, and more responsive. Taking the snow removal operation on an elevated bridge as an example, the entire task can be divided into five sub-segments, corresponding to three snow removal phases: the initial clearing phase, encompassing segments 1 and 2, is used for the initial removal of large areas of snow; the concentrated blowing phase, encompassing segment 3, deals with residual snow in dense areas; and the residual snow drying phase, corresponding to segments 4 and 5, is used for detailed heating and drying.

[0149] The deployment model is as follows:

[0150]

[0151] in,

[0152] : jet power output of the ith control segment;

[0153] : The total number of segments into which the snow removal operation cycle is divided;

[0154] : Control the increase or decrease of each section based on the energy consumption optimization coefficient extracted from historical mission data;

[0155] The model adopts a nonlinear regulation method based on a sine function to achieve multi-segment adaptive power output driven by task subdivision.

[0156] During actual snow removal operations, turbojet snowplows face the challenge of constantly changing workloads. Initially, the focus may be on removing thick snow, while mid-stage operations shift to evenly clearing any remaining snow. Finally, the emphasis is on cleaning, drying, and heat treatment to prevent ice formation. In this dynamic, multi-stage environment, adopting a unified power control standard would make it difficult to meet the varying demands for power output accuracy and intensity at each stage.

[0157] This step establishes a deployment model based on the turbojet engine response mechanism, discretizing the entire operation cycle into multiple control segments on the time axis. Each segment sets a different output strategy based on the stage goals, structural thermal sensitivity, and environmental energy consumption trade-offs, thereby realizing a layer-by-layer power allocation system from coarse adjustment to fine adjustment to fine control.

[0158] S6. Perform interruption and fallback operations on the segmented-controlled turbojet engine power in abnormal situations.

[0159] Traditional turbojet snow removal control systems often rely on fixed threshold strategies for handling exceptions, such as simply triggering an emergency shutdown or alarm when temperature or jet pressure exceeds a set value. While intuitive and easy to implement, this approach suffers from slow response, high false trigger rates, and poor adaptability when dealing with complex combinations of abnormal signals (such as critical changes in multiple boundary indicators). Furthermore, it lacks the ability to dynamically determine the impact of scenarios and historical behavior.

[0160] The turbojet snow blower involved in this step can receive temperature curves, wind pressure fluctuations and abnormal nozzle acoustic data in real time, and perform diagnosis once it is found that the parameters exceed the threshold.

[0161] If a sudden change in wind pressure or an acoustic anomaly at the nozzle exceeds a set threshold, the expert network system is immediately activated for diagnosis. Combining historical failure cases, jet behavior patterns, and material thermal damage characteristics, a rapid assessment is made regarding the presence of structural or thermal overload risks. Diagnostic results are typically categorized as "safe," "critical," or "structural impact exceeding the specified threshold." If a severe anomaly is identified, the system jumps directly to a safe fallback power level, reducing the current power by a factor to prevent further structural or road damage.

[0162] According to the diagnosis results, the current power output Make corrections, that is, continue to execute the current power output; or directly jump to the safe fallback power value and reduce the power based on the fallback coefficient to avoid irreversible damage such as nozzle erosion and road cracking.

[0163] Safety fallback power value The judgment model is as follows:

[0164]

[0165] in,

[0166] : Back-off coefficient, a constant less than 1, represents the output compression ratio after risk control, and is generally set by the expert network system according to the event level.

[0167] After the sensor data (including temperature, pressure, sound waves, etc.) collected by the system in real time enters the expert experience network, the network judges the data based on multiple rule layers, including but not limited to the following criteria:

[0168] If the diagnosis result is "safe state", the current power output is maintained and the ;

[0169] If the diagnosis result is "critical risk" or "structural impact exceeds the standard", the output power will be reduced proportionally, that is, ;

[0170] Backoff coefficient The specific value of is determined by the risk level set in the expert network. For example, mild fallback uses 0.85, moderate is 0.7, and severe may drop below 0.5.

[0171] The control module receives the final After that, the jet intensity is readjusted and the power response correction is completed quickly, thereby avoiding safety hazards such as material overheating, mechanical loss or abnormal structural vibration.

[0172] For example, if the system detects an abnormally high nozzle temperature rise rate, but a stable pressure curve and no fission signals in the acoustic signature, the expert network can determine a "moderate warning" and reduce the output power to 70%. This flexible control strategy avoids energy waste caused by false triggering and restarting the entire machine, ensuring uninterrupted mission completion and significantly improving intelligent resilience and control stability.

[0173] Example 2:

[0174] Based on the power control method for a turbojet snowplow described in Example 1, a surface material identification and adaptation mechanism is introduced during the time-domain segmented control of the turbojet engine power output in step S5. Before starting segmented power control, visual recognition and spectral reflectance analysis are used to determine the type of surface material being used (concrete, high-strength asphalt, or rubber pavement). Based on the material's thermal expansion coefficient and thermal tolerance limit, a corresponding power curve segmentation template is selected. Using preset curve templates from a material classification database, the jet output distribution is further refined, ensuring optimal efficiency for snow removal tasks under different surface materials while ensuring safety. Incorporating the surface material identification mechanism, the power segmentation curve is no longer "one-size-fits-all"; instead, the jet output curve is individually adapted based on the thermal response characteristics of different materials (such as concrete, rubber, and asphalt), significantly reducing the risk of thermal fatigue damage. Compared to traditional control methods that rely solely on time or task phases as the classification criteria, this method demonstrates higher levels of intelligent working condition perception and control precision.

[0175] During the implementation of this mechanism, a camera module installed on the front end of the turbojet snowplow first performs rapid, non-contact identification of the operating surface to determine the type of surface material. Once identification is complete, the system automatically calls upon the built-in material database and selects a matching jet power curve template based on parameters such as the corresponding material's thermal load limit, thermal diffusion coefficient, and structural expansion response. This template includes the location of segment points, power increase / decrease rates, and the maximum allowable power limit, ensuring that each operation meets snow removal efficiency while avoiding permanent structural damage due to heat loss from surface materials. Especially in scenarios with multi-material transition zones (such as asphalt to concrete), the system will dynamically switch jet strategies to achieve continuous operation across multiple materials.

[0176] At the same time, during the actual operation of the system, especially in scenarios such as emergency snow removal, airport emergency cleaning, and critical access security, tasks often face the practical need for rapid recovery after interruption. After step S6, a task priority-based interruption recovery strategy was designed. This strategy does not rely on any complex control models. Based solely on the task level, remaining workload, and time window, it uses a decision logic tree to quickly determine whether to continue execution, wait for recovery, force interruption, or transfer the task, and respond accordingly.

[0177] Multiple task levels are preset (for example: high priority = airport main runway, medium priority = apron, low priority = auxiliary channel). When the external sensor system detects abnormal energy consumption, operation timeout, or extreme low temperature shock that causes the equipment to temporarily shut down, the system will automatically record the current task progress status and queue up to resume operation in priority order when conditions permit, ensuring the continuous execution of critical tasks. By introducing task priority logic, the traditional "interrupt exit when operation is abnormal" strategy is upgraded to a high-reliability scheduling framework that supports task slicing recording and recovery. Especially in airport environments with high-frequency critical tasks and tight resources, it can significantly reduce task losses caused by equipment downtime, improve the continuity and emergency recovery capabilities of snow removal operations, and has extremely high practicality and engineering value.

[0178] Example 3:

[0179] This embodiment also provides a computer device, which is suitable for a power control method of a turbojet snow remover based on artificial intelligence, including a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute computer-executable instructions to implement a power control method of a turbojet snow remover based on artificial intelligence proposed in the above embodiment.

[0180] This embodiment also provides a storage medium having a computer program stored thereon. When the program is executed by a processor, the power control method of a turbojet snow blower based on artificial intelligence proposed in the above embodiment is implemented.

[0181] The computer device may be a terminal, comprising a processor, memory, a communication interface, a display, and an input device connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores an operating system and computer programs. The internal memory provides an environment for the operating system and computer programs stored in the non-volatile storage media. The communication interface of the computer device is used to communicate with external terminals via wired or wireless communication. Wireless communication may be achieved via Wi-Fi, a carrier network, NFC (near-field communication), or other technologies. The display of the computer device may be a liquid crystal display or an electronic ink display. The input device may be a touchscreen overlay on the display, buttons, a trackball, or a touchpad on the computer device housing, or an external keyboard, touchpad, or mouse.

[0182] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the various embodiments of the method of the present invention. The aforementioned storage medium includes various media that can store program code, such as USB flash drives, mobile hard drives, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical disks.

[0183] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0184] More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic devices), a portable computer disk cartridge (magnetic devices), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disc read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting, or processing it in another suitable manner as necessary, and then storing it in a computer memory.

[0185] It should be understood that various parts of the present invention can be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used: a discrete logic circuit having logic gate circuits for implementing logic functions on data signals, an application-specific integrated circuit having suitable combinational logic gate circuits, a programmable gate array (PGA), a field-programmable gate array (FPGA), etc.

[0186] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A power control method for a turbojet snow blower based on artificial intelligence, characterized in that: The following steps are involved: S1. Perform scene type image recognition, use feature vectors to map and compare with semantic label sets, and output the semantic region level identification of the current scene; S2. Map the output semantic region level identifier to its corresponding working environment sensitivity vector to quantify the sensitivity of the region to environmental factors such as temperature and wind speed; S3. Based on the output working environment sensitivity vector, construct a primary power strategy function and generate a preliminary jet power strategy; S4. Based on the preliminary jet power strategy, the power is adaptively adjusted by detecting the residual snow thickness and surface temperature rise in the snow removal area, and the adjusted control strategy is output; S5. Based on the output control strategy, the turbojet engine power output is controlled in a time domain segmented manner; S6. Perform interruption and fallback operations on the segmented-controlled turbojet engine power in abnormal situations.

2. The power control method of a turbojet snow blower based on artificial intelligence according to claim 1, characterized in that: In step S1, real-time image data of the operation scene is collected by the front-end camera module and the geographic tag system of the turbojet snow blower, and feature vectors of the operation area are extracted based on the multi-level convolutional attention network of artificial intelligence, including surface texture, structural boundaries, and facility types; Construct an image scene recognition function, use the feature vector to map and compare with the semantic label set, and output the semantic area level identification value of the current scene.

3. The power control method of a turbojet snow blower based on artificial intelligence according to claim 2, characterized in that: The image scene recognition function is as follows: ； in, : The semantic area level identifier of the current scene, including airport runways, viaducts, and suburban roads; : original image input; : The semantic features extracted from the image by the i-th feature channel; : The importance weight of each semantic feature; : Post-processing mapping function, used to classify the multi-feature combination results into discrete semantic levels; The scene recognition result of the image scene recognition function provides structured prior knowledge for subsequent power adjustment.

4. The power control method of a turbojet snow blower based on artificial intelligence according to claim 1, characterized in that: In step S2, each scene type corresponds to a set of environmental sensitivity features. The semantic region level identifiers are automatically mapped to the work environment sensitivity vector through the parameter database, and an environmental response factor model is constructed. The work environment sensitivity vector is weighted and corrected based not only on static standards but also by introducing current real-time meteorological data. The environmental response factor model is as follows: ； in, : Operating environment sensitivity vector, used to guide subsequent power control; :According to semantic region level identification Preset environment response template vector; :Based on current real-time weather data A correction function for correcting the environmental sensitivity vector; The model builds a logical association bridge between semantic types and real-time physical environments, which is used for continuous mapping from image recognition to environmental constraint modeling.

5. The power control method of a turbojet snow blower based on artificial intelligence according to claim 1, characterized in that: In step S3, after obtaining the operating environment sensitivity vector, a primary power strategy function is generated by setting a logic builder in combination with the response capability of the available jet modules on the turbojet snow blower, so that the jet output can effectively remove snow without causing excessive thermal stress, wind erosion, or local damage to the operating area; The primary power strategy function is as follows: ； in, : Candidate jet power control value, the unit is not fixed, indicating the relative level; : The three main dimensions of the operating environment sensitivity vector include temperature sensitivity, wind speed adaptability, and structural response limit; : Primary power control function based on operating environment sensitivity vector; : System preset weights are set based on device performance and historical experience; The output of this function is Instead of making a final control output, a candidate power strategy that can be optimized is constructed, and adaptive power adjustment is performed in subsequent steps.

6. The power control method of a turbojet snow blower based on artificial intelligence according to claim 1, characterized in that: In step S4, the turbojet snow remover is equipped with a surface feedback sensor module, including a residual snow imaging detector, an infrared temperature sensor, and a micro laser ranging device. The module collects the residual snow thickness, surface humidity, and temperature rise curve of the operating area once per second to determine whether the current jet power reaches the ideal snow removal effect and the thermal impact threshold. Feedback mechanism with candidate jet power control value With the detection data as input and the control condition, a snow state adaptation function is constructed to realize power adaptive adjustment and output the actual jet power value to prevent damage to the ground structure caused by insufficient or excessive power.

7. The power control method of a turbojet snow blower based on artificial intelligence according to claim 6, characterized in that: The snow state adaptation function is as follows: ； in, : The actual jet power value; : The currently detected residual snow thickness; : The lower limit of ideal clearing thickness set by the system; : The current detected surface temperature; : The maximum acceptable surface temperature of the structural material; : Regulation factor, representing the intensity of feedback effect.

8. The power control method of a turbojet snow blower based on artificial intelligence according to claim 1, characterized in that: In step S5, the turbojet engine power output is controlled in time domain segmentation, the total mission cycle is divided into multiple control segments, and the actual jet power value in each segment is controlled. For nonlinear power allocation, the allocation model is as follows: ； in, : jet power output of the ith control segment; : The total number of segments into which the snow removal operation cycle is divided; : Control the increase or decrease of each section based on the energy consumption optimization coefficient extracted from historical mission data; The model adopts a nonlinear regulation method based on a sine function to achieve multi-segment adaptive power output driven by task subdivision.

9. The power control method of a turbojet snow blower based on artificial intelligence according to claim 1, characterized in that: In step S6, the turbojet snow blower receives temperature curves, wind pressure fluctuations, and nozzle abnormal acoustic data in real time, and performs a diagnosis once it is found that the parameters exceed the threshold value; According to the diagnosis results, the current power output Make corrections, that is, continue to execute the current power output; or directly jump to the safe fallback power value to avoid irreversible damage such as nozzle erosion and road cracking; Safety fallback power value The judgment model is as follows: ； in, : Back-off coefficient, a constant less than 1, determined based on the feedback from the expert network.

10. The power control method of a turbojet snow blower based on artificial intelligence according to claim 8, characterized in that: When performing time-domain segmented control of the turbojet engine power output, a surface material identification and adaptation mechanism is also introduced. Before starting the power segmented control, visual recognition and spectral reflectance analysis are used to determine the type of working pavement material, including concrete, high-strength asphalt, and rubber pavement. Based on the material's thermal expansion coefficient and thermal tolerance limit, the corresponding power curve segmentation template is selected.

Images