CN117822371A

CN117822371A - Road anti-icing method, device, equipment and storage medium

Info

Publication number: CN117822371A
Application number: CN202410246519.1A
Authority: CN
Inventors: 魏云波; 武生彪; 鲁浩; 万军; 尹正文; 何成滔; 林梦; 刘若慨; 杨雄兵; 赵兴宗
Original assignee: PowerChina Kunming Engineering Corp Ltd
Current assignee: PowerChina Kunming Engineering Corp Ltd
Priority date: 2024-03-05
Filing date: 2024-03-05
Publication date: 2024-04-05

Abstract

The application discloses an anti-icing method, device, equipment and storage medium for roads, which relate to the technical field of electric digital data processing. The application has utilized throwing simplicity, the homogeneity of solution, has replaced directly throwing sodium chloride solid for the road can be evenly covered by the solution, and the anti-icing method of this application has been ensured to have no hysteresis quality to above characteristics and advantage, can not be redundant for salt.

Description

Road anti-icing method, device, equipment and storage medium

Technical Field

The present disclosure relates to the field of digital data processing technologies, and in particular, to a road anti-icing method, apparatus, device, and storage medium.

Background

Road icing is a phenomenon which is easy to generate in cold weather, if the ground temperature is lower than 0 ℃, snow accumulation or icing can occur on the road, so that the friction coefficient of the road is greatly reduced, the traveling of pedestrians and vehicles is influenced, and even the safety of the pedestrians and the vehicles is influenced.

The treatment measures for road icing are to install warning signs such as road icing on road sections easy to freeze, seal roads which cannot pass safely, and throw industrial salt on the road sections which are sealed and cannot be recovered for a long time through special vehicles so as to accelerate the thawing speed of icing.

At present, the treatment measures for the ice-prone road section are lagged, the operation of sealing or spraying salt is usually adopted when the ice-prone road section is seriously passed, and a large amount of salt is sprayed to the ice-prone road section by a means of redundant salt feeding for special vehicles, so that excessive salt can be passively carried to other places by rainwater, pedestrians and vehicles in subsequent passing, such as agriculture, animal husbandry, drinking water and the like near the road, pollution is caused, or chlorine ions in the salt have coordination effect on most metals and form complex metals after coordination, thereby forming corrosion and causing property loss.

Disclosure of Invention

The main object of the present application is to provide an anti-icing method, device, equipment and storage medium for roads, so as to solve the problems of hysteresis and redundant processing of processing measures for ice-prone road sections in the prior art.

In order to achieve the above purpose, the present application provides the following technical solutions:

an anti-icing method for a road, wherein one side or two sides of the road are fixedly provided with spraying devices for spraying sodium chloride solution with preset concentration, the spraying devices have a solution preparation function, and fog outlets of all the spraying devices face the road, and the anti-icing method comprises the following steps:

acquiring a plurality of historical environmental temperature curves of the road;

training and learning all historical environmental temperature curves through a machine learning algorithm to obtain an environmental temperature prediction model;

predicting the future environmental temperature of the road through the environmental temperature prediction model, and generating a future environmental temperature curve;

acquiring a curve of which the temperature is less than or equal to zero ℃ in the future environmental temperature curve, and marking the curve as a water ice-forming curve;

respectively obtaining the minimum value and the initial time stamp of the water ice-making curve;

obtaining the concentration of the sodium chloride solution matched with the minimum value so that the freezing point of the sodium chloride solution is equal to the minimum value;

Adding a preset fault tolerance constant to the concentration of the sodium chloride solution to obtain the concentration of the sodium chloride solution to be prepared;

completing preparation in the spraying device according to the concentration of the sodium chloride solution to be prepared to obtain a prepared sodium chloride solution;

and atomizing the prepared sodium chloride solution through the mist outlet based on the starting time stamp and spraying the atomized sodium chloride solution to the road at a preset speed.

As a further improvement of the present application, the prepared sodium chloride solution is atomized through the mist outlet based on the start time stamp and sprayed to the road at a preset rate, and thereafter, comprises:

detecting whether a target object exists in a preset range of each spraying device or not through a target detection algorithm;

if the target object exists, closing a spraying device with the target object in a preset range;

acquiring the closing time of the spraying device, and judging whether the closing time is greater than or equal to a preset time;

if yes, a manual intervention request is generated and sent to an external receiving end.

As a further improvement of the present application, detecting whether a target object exists in a preset range of each spraying device by a target detection algorithm includes:

Acquiring image data of the preset range based on a current spraying device;

dividing the image data into a plurality of grids through grids with preset density;

predicting a preset number of first-order bounding boxes based on each grid respectively, wherein each first-order bounding box comprises at least one grid;

defining that the target object has the highest confidence;

respectively acquiring the confidence coefficient of each first-order bounding box;

acquiring a first-order boundary box with highest confidence coefficient from all confidence coefficients as a second-order boundary box;

calculating the intersection ratio of the second-order boundary frame and the first-order boundary frame respectively;

reserving a second-order boundary box with the cross ratio being greater than or equal to a preset ratio as a third-order boundary box;

the third-order boundary frame with the highest confidence coefficient is reserved and used as a fourth-order boundary frame;

acquiring a union set of all fourth-order boundary frames, wherein the union set is a detection frame of the target object;

judging whether the preset range is provided with the detection frame or not.

As a further improvement of the present application, training and learning all historical ambient temperature curves by a machine learning algorithm to obtain an ambient temperature prediction model includes:

generating a plurality of acquisition points based on a preset time interval;

respectively acquiring each historical environmental temperature curve based on all acquisition points, and generating a data set based on each historical environmental temperature curve;

Integrating all the data sets to form a total data set;

normalizing the total data set to obtain a normalized data set;

dividing the normalized data set into a training set and a verification set according to a preset proportion;

defining a topological relation of a neural network model, wherein the topological relation comprises an input layer, an implicit layer and an output layer which are sequentially connected through signals;

outputting the training set to the input layer, carrying out iteration for preset times through the neural network model, and respectively acquiring root mean square errors of the training set and the current training result based on each iteration;

and acquiring the minimum value in all root mean square errors, and acquiring a training result corresponding to the minimum value as the environmental temperature prediction model.

As a further improvement of the present application, the topological relation is characterized by formula (1):

（1）；

wherein,for the godThrough a network model; />Is the +.>A plurality of input nodes, each input node corresponding to a set of training data of the training set,/->Is the +.>The input node is to the ++>Preset weights of the input nodes; />For connecting to the +. >A threshold of the input nodes;is a transfer function, and->。

the mass of sodium chloride-containing solid of the sprayed sodium chloride solution is obtained;

acquiring the evaporation mass of the sprayed sodium chloride solution based on the water ice-forming curve through a fixed integral to obtain the solution residual mass of the sprayed sodium chloride solution;

calculating the concentration of the sodium chloride solution on the road according to the mass of the sodium chloride-containing solid and the residual mass of the solution;

defining a dynamic proportionality coefficient based on the preset rate, and linearly reducing the preset rate through the dynamic proportionality coefficient;

and searching the optimal value of the dynamic proportionality coefficient through a global optimizing algorithm so that the concentration of the sodium chloride solution on the road is equal to the saturation value concentration of the current environment temperature.

As a further improvement of the present application, searching for an optimal value of the dynamic scaling factor by a global optimization algorithm so that the concentration of the sodium chloride solution on the road surface is equal to the saturation value concentration of the current environmental temperature, including:

According to the formula (2), a plurality of random solutions are given to the optimal value, and the result of defining all random solutions is that the concentration of the sodium chloride solution of the pavement is equal to the saturation value concentration of the current environmental temperature;

（2）；

wherein,for the set of all random solutions, +.>For each random solution, respectively->Label for random solution->The number of all random solutions; />For the set of velocities for all the random solutions,the speed of each random solution;

initializing the position of each random solution, and respectively updating the current position and the current speed according to the formula (3) based on the same random solution:

（3）；

wherein,is->The random solution is at->Speed of walking->Is->The random solution is at->Speed inertia of steps,/->Is an inertia coefficient>Is->Self-cognition characterization of the individual random solutions,is->Social cognitive characterization of individual random solutions; />And->Are all the learning factors of the human body,for interval->Random number of->Is->Individual optimal solutions, which have been obtained for the individual random solutions, < >>Globally optimal solution, which has been obtained for all random solutions, < >>Is->Step->A random solution;

respectively iterating each random solution for a preset number of times according to the formula (3), and linearly decrementing the inertia coefficient by the formula (4) once at each iteration to update each Said->；

（4）；

Wherein,is->The random solution is at->After step optimizationInertia coefficient->For initial inertia factor, +.>For the current iteration step +.>The maximum iteration step number;

respectively judge eachComparing whether the first difference value of the previous iteration is smaller than or equal to a first preset adaptation threshold value;

if each isComparing with the first difference value of the previous iteration being smaller than or equal to a first preset adaptation threshold value, respectively judging the +.>Comparing whether the second difference value of the previous iteration is smaller than or equal to a second preset adaptation threshold value;

if saidAnd comparing the second difference value of the previous iteration with a second preset adaptation threshold value, and judging that the optimal solution of the optimal value is obtained.

In order to achieve the above purpose, the present application further provides the following technical solutions:

an anti-icing device for a roadway, the anti-icing device being applied to an anti-icing method for a roadway as described above, the anti-icing device comprising:

the historical environment temperature curve acquisition module is used for acquiring a plurality of historical environment temperature curves of the road;

the environment temperature prediction model acquisition module is used for training and learning all historical environment temperature curves through a machine learning algorithm to obtain an environment temperature prediction model;

The future environmental temperature curve acquisition module is used for predicting the future environmental temperature of the road through the environmental temperature prediction model and generating a future environmental temperature curve;

the water icing curve acquisition module is used for acquiring a curve with the temperature less than or equal to zero ℃ in the future environmental temperature curve and marking the curve as a water icing curve;

the water ice-making curve characteristic acquisition module is used for respectively acquiring the minimum value and the initial time stamp of the water ice-making curve;

the sodium chloride solution concentration acquisition module is used for acquiring the concentration of the sodium chloride solution matched with the minimum value so as to enable the freezing point of the sodium chloride solution to be equal to the minimum value;

the preset fault tolerance constant adding module is used for adding a preset fault tolerance constant to the concentration of the sodium chloride solution to obtain the concentration of the sodium chloride solution to be prepared;

the sodium chloride solution preparation module is used for completing preparation in the spraying device according to the concentration of the sodium chloride solution to be prepared to obtain a prepared sodium chloride solution;

and the sodium chloride solution output control module is used for atomizing the prepared sodium chloride solution through the mist outlet based on the starting time stamp and spraying the atomized sodium chloride solution to the road at a preset speed.

an electronic device comprising a processor, a memory coupled to the processor, the memory storing program instructions executable by the processor; the processor, when executing the program instructions stored in the memory, implements an anti-icing method for a roadway as described above.

a storage medium having stored therein program instructions which when executed by a processor implement an anti-icing method enabling a road as described above.

The method comprises the steps of obtaining a plurality of historical environmental temperature curves of a road; training and learning all historical environmental temperature curves through a machine learning algorithm to obtain an environmental temperature prediction model; predicting the future environmental temperature of the road through an environmental temperature prediction model, and generating a future environmental temperature curve; acquiring a curve of which the future environmental temperature curve is less than or equal to zero ℃ and marking the curve as a water ice-forming curve; respectively obtaining the minimum value and the initial time stamp of the water icing curve; obtaining the concentration of the sodium chloride solution matched with the minimum value so that the freezing point of the sodium chloride solution is equal to the minimum value; adding a preset fault tolerance constant to the concentration of the sodium chloride solution to obtain the concentration of the sodium chloride solution to be prepared; completing preparation in a spraying device according to the concentration of the sodium chloride solution to be prepared to obtain a prepared sodium chloride solution; the formulated sodium chloride solution was atomized through the mist outlet based on the starting time stamp and sprayed onto the road at a preset rate. According to the method, a plurality of historical data are learned and trained through machine learning, an icing prediction model of the road section is obtained, information such as an icing time starting point and the lowest temperature of the road section is predicted through the prediction model, so that the concentration of sodium chloride solution required for preventing icing at the lowest temperature is calculated, the sodium chloride solution with the concentration is prepared through a spraying device with a solution preparation function (the device is the prior art), and meanwhile, the solution with the corresponding concentration is thrown at the stage that the icing action of a road just starts to occur or is prepared to occur, and therefore the icing action cannot start. The application utilizes the throwing simplicity and uniformity of the solution to replace the direct throwing of sodium chloride solid, so that the road can be uniformly covered by the solution. The above features and advantages ensure that the anti-icing method of the present application has no hysteresis and is not redundant to salt.

Drawings

FIG. 1 is a schematic flow chart of steps of an embodiment of an anti-icing method for a roadway according to the present application;

FIG. 2 is a schematic view of an embodiment of an anti-icing device for a roadway according to the present application;

FIG. 3 is a schematic structural diagram of one embodiment of an electronic device of the present application;

FIG. 4 is a schematic diagram illustrating the structure of one embodiment of a storage medium of the present application.

Detailed Description

The following description of the technical solutions in the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings in the embodiments of the present application, and it is apparent that the described embodiments are only some embodiments of the present application, not all embodiments. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are within the scope of the present disclosure.

The terms "first," "second," "third," and the like in this application are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "first," "second," and "third" may explicitly or implicitly include at least one such feature. In the description of the present application, the meaning of "plurality" means at least two, for example, two, three, etc., unless specifically defined otherwise. All directional indications (such as up, down, left, right, front, back … …) in the embodiments of the present application are merely used to explain the relative positional relationship, movement, etc. between the components in a particular gesture (as shown in the drawings), and if the particular gesture changes, the directional indication changes accordingly. Furthermore, the terms "comprise" and "have," as well as any variations thereof, are intended to cover a non-exclusive inclusion. For example, a process, method, system, article, or apparatus that comprises a list of steps or elements is not limited to only those listed steps or elements but may include other steps or elements not listed or inherent to such process, method, article, or apparatus.

Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the present application. The appearances of such phrases in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. Those of skill in the art will explicitly and implicitly appreciate that the embodiments described herein may be combined with other embodiments.

As shown in fig. 1, the present embodiment provides an embodiment of an anti-icing method for a road, in the present embodiment, a spraying device for spraying a sodium chloride solution with a preset concentration is fixedly installed on one side or both sides of the road, the spraying device has a solution preparation function, and mist outlets of all the spraying devices face the road.

Preferably, the spraying device is a mature prior art, in which a plurality of types of spraying devices with solution preparation functions meet the use requirements of the embodiment (for example, "202110265878.8-a solution preparation method and solution preparation device," "202122392684.2 solution preparation device," and other items of a large number of solution preparation devices), the embodiment simply lists only core components, and details of the devices are not repeated and shown in the drawings, so that in order to realize the spraying functions of the devices, only an atomizing element is required to be installed at a solution outlet, and the spraying device is the same as the prior art.

Preferably, the spraying device with the solution preparation function comprises a box body fixedly arranged on one side or two sides of a road, a solute carrier (such as a bottle, a box and other commonly-contained containers for containing sodium chloride) and a solvent carrier (such as a bottle, a barrel, a cylinder and other commonly-used containers for containing water), a solid metering piece with one end communicated with the solute carrier (used for obtaining accurate solute mass before solution preparation), and a liquid metering piece with one end communicated with the solvent carrier (used for obtaining accurate solvent mass before solution preparation), wherein the other end of the solid metering piece is communicated with the other end of the liquid metering piece through a solution preparation bin.

Preferably, after metering is completed, the solid metering element releases the solute into the preparation bin, the liquid metering element releases the solute into the preparation bin, and a stirring element is arranged in the preparation bin to fully mix the solute and the solvent, so that the sodium chloride solution meeting the requirements of the embodiment is formed.

Preferably, the configuration bin is communicated with the atomizing element through a pipeline, and is conveyed to the atomizing element through the pipeline after the sodium chloride solution is prepared and sprayed to an external road.

Preferably, in order to prevent the solvent in the solvent carrier from freezing along with the external ambient temperature, a heating and insulating member can be installed on the solvent carrier (only water needs to be prevented from freezing, and the temperature after heating is higher than the freezing point, so that the operation and maintenance cost is reduced).

It should be noted that, the above devices and structures thereof are all in the prior art, the present embodiment also does not creatively improve the above devices and structures thereof, and is a conventional application of the above devices and structures thereof, and the present embodiment aims to provide a method instead of providing a device, so that the present embodiment does not need to describe the device and show the drawings in detail.

Specifically, the anti-icing method of the present embodiment includes the steps of:

step S1, a plurality of historical environmental temperature curves of a road are obtained.

Preferably, the historical ambient temperature profile can be obtained directly by measurement or by the weather bureau.

And S2, training and learning all historical environmental temperature curves through a machine learning algorithm to obtain an environmental temperature prediction model.

Preferably, the machine learning algorithm includes decision tree, random forest, K-nearest neighbor, naive bayes, support vector machine, neural network, etc., and the machine learning algorithm of this embodiment adopts bp neural network.

And S3, predicting the future environmental temperature of the road through an environmental temperature prediction model, and generating a future environmental temperature curve.

Preferably, the historical ambient temperature curve and the future ambient temperature curve may be based on the same horizontal axis attribute, i.e. natural time, and the unit scale (which may be understood as the acquisition interval) may be set to 1min, 3min, 5min, 10min, etc., in order to ensure timeliness of the method of the embodiment, and prevent hysteresis, it is recommended not to exceed 15min.

And S4, acquiring a curve with the temperature less than or equal to zero ℃ in a future environmental temperature curve, and marking the curve as a water ice-forming curve.

Preferably, the zero point of the vertical axis of the water ice-making curve is the freezing point of water.

And S5, respectively obtaining the minimum value and the starting time stamp of the water icing curve.

Preferably, the minimum value and the start time stamp are directly read from the curve and the horizontal and vertical axes of the coordinate system in which the curve is located.

And S6, acquiring the concentration of the sodium chloride solution matched with the minimum value so that the freezing point of the sodium chloride solution is equal to the minimum value.

Preferably, there is a formula for calculating the freezing point of the solution according to Raoult's law and freezing point depression:

wherein->Degree celsius reduced for the freezing point of the solution (water), ->Is the molar freezing point depression constant, +.>Is the molar concentration of solute (sodium chloride).

Preferably, the molar freezing point depression constant is a characteristic parameter inherent to the substance that is indicative of the depression of the freezing point caused by a unit molar concentration of solute in the solvent. Different solvents have different freeze point depression constants, and the same solvent may have different freeze point depression constants at different temperatures. The freezing point depression constants of commonly used solvents can be obtained directly from the relevant physicochemical handbook.

And S7, adding a preset fault tolerance constant to the concentration of the sodium chloride solution to obtain the concentration of the sodium chloride solution to be prepared.

Preferably, the sodium chloride solution concentration can be characterized by a percent concentration or a molar concentration, and the units of the preset fault tolerance constant are unified with the units of the sodium chloride solution concentration. If a percentage concentration is employed, the preset fault tolerance constant may be set to any constant between 0.5% and 1%; if a molar concentration is employed, the preset fault tolerance constant may be set to any constant between 0.05mol/L and 0.1 mol/L.

It should be noted that the preset fault tolerance constant needs to be added to the sodium chloride solution concentration rather than subtracted.

And S8, completing preparation in a spraying device according to the concentration of the sodium chloride solution to be prepared, and obtaining the prepared sodium chloride solution.

Preferably, the preparation method of the spraying device refers to the structure and principle of the spraying device in the prior art, and this embodiment is not repeated.

And step S9, atomizing the prepared sodium chloride solution through a fog outlet based on the initial timestamp and spraying the atomized sodium chloride solution to a road at a preset speed.

Preferably, the spray coverage surface may be provided as a sector, a plurality of spray devices being arranged alternately or side by side so that all sectors cover the road.

Further, step S9, atomizing the prepared sodium chloride solution through the mist outlet based on the start time stamp and spraying the atomized sodium chloride solution onto the road at a preset rate, and then further comprises the steps of:

step S10, detecting whether a target object exists in a preset range of each spraying device or not through a target detection algorithm.

Preferably, the YOLO algorithm detects and acquires whether a target object exists within a preset range of each spraying device.

Step S20, if the target object exists, closing the spraying device with the target object in the preset range.

Step S30, acquiring the closing time of the spraying device, and judging whether the closing time is greater than or equal to the preset time.

And S40, if the closing time length is greater than or equal to the preset time length, generating a manual intervention request and sending the manual intervention request to an external receiving end.

Further, step S10, detecting whether a target object exists in a preset range of each spraying device through a target detection algorithm, specifically includes the following steps:

step S101, acquiring image data of a preset range based on the current spraying device.

In step S102, the image data is divided into a plurality of grids by grids of a preset density.

Preferably, the number of grids may be set to 448×448, and the size of the image data may be adjusted to meet the specification of 448×448, and the resized picture may be equally divided into s×s (e.g., 7×7) grids, with the size of each grid being 64×64.

Step S103, predicting a preset number of first-order bounding boxes based on each grid respectively, wherein each first-order bounding box comprises at least one grid.

Step S104, defining the target object with the highest confidence.

Preferably, the target object includes an object that a pedestrian, a vehicle, an animal, or the like will move and possibly use the road.

Step S105, the confidence coefficient of each first-order bounding box is obtained respectively.

Step S106, a first-order boundary box with highest confidence is obtained from all the confidence coefficients and is used as a second-order boundary box.

In step S107, the intersection ratios of the second-order bounding boxes and the first-order bounding boxes are calculated.

Preferably, the intersection ratio is the intersection of the second order bounding box with each first order bounding box, respectively, divided by the union of the second order bounding box with each first order bounding box, respectively, to obtain the ratio (which may be an area ratio).

Step S108, reserving a second-order boundary box with the cross ratio being greater than or equal to a preset ratio as a third-order boundary box.

Step S109, reserving a third-order boundary box with highest confidence as a fourth-order boundary box.

Step S110, a union set of all four-order boundary frames is obtained, and the union set is the detection frame of the target object.

Step S111, judging whether the preset range has a detection frame.

Preferably, each grid is used for predictionCoordinates and width and height of the three-order bounding boxes,and the confidence level of each third-order bounding box, i.e. each grid needs to predict +.>A value.

It will be appreciated that each grid requires predictionPersonal->The method comprises the steps of carrying out a first treatment on the surface of the Wherein (1)>For the offset of the center of the third order bounding box relative to the grid, +.>For the ratio of the third-order bounding box relative to the resized picture, +.>The confidence of the grid is 1 or 0.

Preferably, confidence may be understood as the accuracy of whether there is a target within the current grid and a third order bounding box.

Illustrating: setting a target in the picture after the size adjustment, and setting the width and the height of the picture after the size adjustment asThen:

dividing the picture into 7×7 grids on average, wherein there is one grid located at the center of the target, and the coordinates of the grid areLet the coordinates of the center of the target be +.>Then can be according to->Calculating the offset:

Preferably, in the actual detection, if the predicted third-order bounding box and the actual bounding box overlap perfectly, the value of the overlap ratio is 1. In the practical application process, the value of the preset ratio can be set to 0.5 to determine whether the predicted bounding box is correct, and the more accurate the bounding box is, the more positively correlated the cross ratio is.

Preferably, the YOLO algorithm also requires training of third-order bounding boxes to improve accuracy of target detection.

And next, training the training model through a preset target training set, and iteratively adjusting the weight and bias of the training model by a certain number of times through a back propagation algorithm so as to reduce the value of a loss function of the training model.

Preferably, the loss function is

。

Wherein,is->+.>Whether the three-order bounding boxes are responsible for the indication function of the target or not is judged to be 1 or 0; />、/>、/>、/>、/>Respectively correspond toFirst->Personal->Predicted values.

It is understood that the loss function includes a coordinate value deviation of the third-order bounding box, a deviation of the confidence, a deviation of the prediction probability (or a class deviation).

Wherein,for the mid-point loss of the third-order bounding box in coordinate value deviation,/->Is the third-order boundary frame width and height loss in coordinate value deviation, >For the deviation of the confidence level, +.>To predict deviations in probability (or class deviations).

It should be noted that, since each grid does not necessarily contain an object, if there is no object in the grid, this will result inThe value of (2) is 0, so that the gradient span in the subsequent back propagation algorithm is too large, so +.>To control the loss of the predicted position of the third-order bounding box and to introduce +.>There is no loss of targets within the control single grid.

Further, step S2, training and learning all the historical environmental temperature curves through a machine learning algorithm to obtain an environmental temperature prediction model, specifically includes the following steps:

step S21, generating a plurality of acquisition points based on a preset time interval.

Preferably, the preset time interval may be 1min, 3min, 5min, 10min, etc. as described above, and in order to ensure timeliness of the method of the embodiment, hysteresis is prevented, it is recommended that the time interval is not more than 15min.

Step S22, each historical environment temperature curve is acquired based on all the acquisition points, and a data set is generated based on each historical environment temperature curve.

In step S23, all the data sets are integrated to form a total data set.

And step S24, carrying out normalization processing on the total data set to obtain a normalized data set.

Preferably, the normalization method of zero-mean normalization (Z-score normalization) is preferred in this embodiment, and this method gives the mean (mean) and standard deviation (standard deviation) of the raw data to normalize the data, and the processed data conforms to the standard normal distribution, that is, the mean is 0 and the standard deviation is 1. For the normalization method, in this embodiment, batch normalization (Batch Normalization) may be used, compared with simple normalization when training is performed on the previous neural network, only normalization is performed on the input layer data, but no normalization is performed on the middle layer, although normalization is performed on the data set of the input node, the data distribution of the input data after matrix multiplication is more likely to be changed greatly, and as the number of network layers of the hidden layer is deepened continuously, the change of the data distribution will be larger and larger, so that the normalization is performed on the middle layer of the neural network by batch normalization, and the training effect is better.

Step S25, dividing the normalized data set into a training set and a verification set according to a preset proportion.

Preferably, the preset ratio is typically 80%: the 20% ratio divides the image data into a training set and a verification set, namely 80% of the data is the training set and 20% of the data is the verification set.

Step S26, defining a topological relation of the neural network model, wherein the topological relation comprises an input layer, an implicit layer and an output layer which are sequentially connected through signals.

Step S27, outputting the training set to an input layer, and carrying out iteration for preset times through a neural network model, and respectively obtaining root mean square errors of the verification set and the current training result based on each iteration.

Preferably, the root mean square error is. Wherein (1)>Is root mean square error>For the number of training sets, +.>Is->True value of the individual training set, +.>Is->Training results after the training of the training sets are completed.

It should be noted that the root mean square error formula is only used for principle illustration, and the letter symbols and meanings of the root mean square error formula are not mutually communicated with other formulas in the embodiment.

And S28, acquiring the minimum value in all root mean square errors, and acquiring a training result corresponding to the minimum value as an environment temperature prediction model.

Further, the topological relation is characterized by the formula (1):

（1）。

wherein,is a neural network model；/>Is the->Each input node corresponds to one group of training data of the training set, ++>Is the->Input node to hidden layer +.>Preset weights of the input nodes; / >To connect to the +.>A threshold of the input nodes; />Is a transfer function, and。

preferably, the number in brackets of the symbol corner mark in formula (1) of this embodiment is the number of layers, e.gThe upper corner mark (2) in (a) is a second layer, namely an implicit layer, < + >>The upper corner marks (1, 2) of (a) are from the first layer to the second layer, namely from the input layer to the hidden layer.

The sign meaning of the transfer function is not communicated with other places.

Preferably, training a model to train a neural network typically requires providing a large amount of data, i.e., a data set; the data sets are generally divided into three categories, namely training set (training set), validation set (validation set) and test set (test set) as described above.

Wherein, one epoch (increase in number) is a process equal to one training with all samples in the training set, which means one forward propagation (forward pass) and one backward propagation (back pass); when the number of samples (i.e., training sets) of one epoch is too large, excessive time may be consumed for performing one training, and it is not necessary to use all data of the training set for each training, the whole training set needs to be divided into a plurality of small blocks, that is, a plurality of batches for performing the training; one epoch is made up of one or more latches, which are part of a training set, with only a portion of the data being used for each training process, i.e., one latch, and one iteration being used for training one latch.

Preferably, the neural network training specifically comprises a Perceptron (Perceptron) composed of two layers of neurons, an input layer receiving external input signals and transmitting to an output layer, wherein the output layer is M-P neurons, and a sensor is arrangedAs a step function, and given a training data set, weight +.>(/>=1, 2,..n), and training threshold +.>Can be obtained by learning->It can be understood that a weight corresponding to a fixed value with a fixed input of-1, 0 +.>。

It should be noted that the step function is not interconnected with the symbolic meaning of the other formulas in the embodiment, and the step function is only schematically illustrated and does not participate in the calculation of the other formulas.

Preferably, the number of times of training the neural network in this embodiment may be set to 500 times.

Preferably, the learning rate of 1 st to 250 th epochs may be set to 0.01, the learning rate of 251 st to 325 th epochs may be set to 0.001, and the learning rate of 326 th to 1000 th epochs may be set to 0.0001.

It can be understood that the neural network training of this embodiment mainly includes the following ideas:

(1) initializing weight and bias items in a network, and initializing parameter values (the weight of an output unit, the bias items, the weight of a hidden unit and the bias items are all parameters of a model) to obtain the output value of each layer of elements for activating forward propagation, thereby obtaining the value of a loss function.

(2) And activating forward propagation to obtain the output value of each layer and the expected value of the loss function of each layer.

(3) According to the loss function, calculating an error term of the output unit and an error term of the hidden unit, calculating various errors, calculating a gradient of a parameter with respect to the loss function or calculating a partial derivative according to a calculus chain law. Solving partial derivatives for vectors or matrixes in the composite function, wherein the editing derivatives of the internal functions of the composite function are always multiplied left; for scalar bias derivative in the composite function, the derivative of the internal function of the composite function can be multiplied left or right.

(4) The weights and bias terms in the neural network are updated.

(5) Repeating the steps (2) - (4) until the loss function is smaller than a preset threshold or the iteration times are used up, and outputting the parameters at the moment to obtain the current optimal parameters.

and step S100, obtaining the mass of sodium chloride solids contained in the sprayed sodium chloride solution.

And step S200, acquiring the evaporation mass of the sprayed sodium chloride solution based on the water ice formation curve through definite integration, and obtaining the solution residual mass of the sprayed sodium chloride solution.

And step S300, calculating the concentration of the sodium chloride solution on the road by the mass of the solid containing sodium chloride and the residual mass of the solution.

Step S400, defining a dynamic proportionality coefficient based on the preset rate, and linearly reducing the preset rate through the dynamic proportionality coefficient.

And S500, searching an optimal value of the dynamic proportionality coefficient through a global optimizing algorithm so that the concentration of the sodium chloride solution on the road is equal to the saturation value concentration of the current environment temperature.

Preferably, the design intent of steps S100 to S500 of this embodiment is that since the sprayed sodium chloride solution itself will not freeze, it must be accompanied by evaporation, resulting in a linear increase in the duty cycle of the solute in the solution, against which a dynamic scaling factor is introduced. It can be understood that the dynamic scaling factor and the linear lifting factor are reciprocal, and the dynamic scaling factor is difficult to directly obtain because the change of the linear lifting factor has various external factors involved, such as wind speed, temperature, direct interference (other substances attached to the moving object are taken away, impurities attached to the moving object are added into the solution), indirect interference (rainwash and loss caused by natural flow), so that the global optimization algorithm is adopted to find the reciprocal value capable of resisting the linear lifting.

Further, step S500 is to find an optimal value of the dynamic scaling factor by the global optimization algorithm so that the concentration of the sodium chloride solution on the road is equal to the saturation value concentration of the current environmental temperature, and specifically includes the following steps:

in step S5001, a plurality of random solutions are given to the optimal value according to the formula (2), and the result of defining all the random solutions is that the concentration of the sodium chloride solution on the road surface is equal to the saturation value concentration of the current ambient temperature.

（2）。

Wherein,for the set of all random solutions, +.>For each random solution, respectively->Label for random solution->The number of all random solutions; />For the set of velocities for all the random solutions,the velocity of each random solution is separate.

Step S5002, initializing the position of each random solution, and updating the current position and the current speed according to equation (3) based on the same random solution:

（3）。

wherein,is->The random solution is at->Speed of walking->Is->The random solution is at->Speed inertia of steps,/->Is an inertia coefficient>Is->Self-cognition characterization of the individual random solutions,is->Social cognitive characterization of individual random solutions; />And->Are all the learning factors of the human body,for interval->Random number of->Is->Individual optimal solutions, which have been obtained for the individual random solutions, < > >Globally optimal solution, which has been obtained for all random solutions, < >>Is->Step (1)/>And (5) randomly solving.

Preferably, the method comprises the steps of,the value range of (2) is +.>Preferably->；/>The value range of (2) is +.>Preferably->。

Step S5003, respectively iterating each random solution for a preset number of times according to formula (3), and linearly decrementing the inertia coefficient by formula (4) once at each iteration to update eachAnd +.>。

（4）。

Wherein,is->The random solution is at->Inertia coefficient after step optimization, ++>For initial inertia factor, +.>For the current iteration step +.>Is the maximum number of iterative steps.

Step S5004 of judging eachWhether the first difference value compared with the previous iteration is smaller than or equal to a first preset adaptation threshold value.

Step S5005, if eachComparing with the first difference value of the previous iteration being smaller than or equal to the first preset adaptation threshold value, respectively judging +.>Whether the second difference value compared with the previous iteration is smaller than or equal to a second preset adaptation threshold value.

Step S5006, ifAnd comparing the second difference value of the previous iteration with a second preset adaptation threshold value, and judging that the optimal solution with the optimal value is obtained.

Preferably, the first preset adaptive threshold and the second preset adaptive threshold can determine the attraction force of the current position to the particle through an adaptive function, and the larger the adaptive function value is, the better the position is, the larger the attraction force to the particle is, and the more likely the adaptive function value is an optimal solution.

Preferably, the fitness function may be one of Griewank function, rastigin function, schaffer function, ackley function, rosenblock function, where the foregoing functions are all typical fitness functions of a global optimization algorithm, and are in the prior art, and specific expressions thereof are not repeated in this embodiment.

In popular sense, the system initializes a group of particle population, each particle has an initial position and speed, calculates an fitness function value according to the fitness function, marks the fitness function value as a current individual history optimal solution, then obtains a global history optimal solution among the particles through cooperative sharing information, iterates to obtain a new speed and position, calculates a new fitness function value of the particle, if the new fitness function value of the particle is greater than the previous individual history optimal solution, the new fitness function value becomes the new individual history optimal solution of the particle, updates the global history optimal solution after the individual history optimal solution of all the particles is updated, and then continues iterating to obtain a new speed and position.

The method comprises the steps of obtaining a plurality of historical environmental temperature curves of a road; training and learning all historical environmental temperature curves through a machine learning algorithm to obtain an environmental temperature prediction model; predicting the future environmental temperature of the road through an environmental temperature prediction model, and generating a future environmental temperature curve; acquiring a curve of which the future environmental temperature curve is less than or equal to zero ℃ and marking the curve as a water ice-forming curve; respectively obtaining the minimum value and the initial time stamp of the water icing curve; obtaining the concentration of the sodium chloride solution matched with the minimum value so that the freezing point of the sodium chloride solution is equal to the minimum value; adding a preset fault tolerance constant to the concentration of the sodium chloride solution to obtain the concentration of the sodium chloride solution to be prepared; completing preparation in a spraying device according to the concentration of the sodium chloride solution to be prepared to obtain a prepared sodium chloride solution; the formulated sodium chloride solution was atomized through the mist outlet based on the starting time stamp and sprayed onto the road at a preset rate. According to the method, a plurality of historical data are learned and trained through machine learning to obtain an icing prediction model of the road section, the information such as the starting point of icing time and the lowest temperature of the road section is predicted through the prediction model, so that the concentration of sodium chloride solution required for preventing icing at the lowest temperature is calculated, then the sodium chloride solution with the concentration is prepared through a spraying device with a solution preparation function (the device is the prior art) and is sprayed at the same time, and the solution with the corresponding concentration is sprayed at the stage when the icing action of the road just begins to occur or is prepared to occur, so that the icing action cannot begin. The embodiment utilizes the throwing simplicity and uniformity of the solution to replace the direct throwing of the sodium chloride solid, so that the road can be uniformly covered by the solution. The above features and advantages ensure that the anti-icing method of the present embodiment has no hysteresis and is not redundant to salt.

As shown in fig. 2, the present embodiment provides a functional module embodiment of an anti-icing device for a road, which is applied to the anti-icing method in the embodiment of the foregoing manner, and the anti-icing device includes a historical ambient temperature curve acquisition module 1, an ambient temperature prediction model acquisition module 2, a future ambient temperature curve acquisition module 3, a water ice curve acquisition module 4, a water ice curve characteristic acquisition module 5, a sodium chloride solution concentration acquisition module 6, a preset fault tolerance constant addition module 7, a sodium chloride solution preparation module 8, and a sodium chloride solution output control module 9 that are electrically connected in sequence.

The historical environment temperature curve acquisition module 1 is used for acquiring a plurality of historical environment temperature curves of a road; the environmental temperature prediction model acquisition module 2 is used for training and learning all historical environmental temperature curves through a machine learning algorithm to obtain an environmental temperature prediction model; the future environmental temperature curve acquisition module 3 is used for predicting the future environmental temperature of the road through an environmental temperature prediction model and generating a future environmental temperature curve; the water icing curve acquisition module 4 is used for acquiring a curve with the temperature less than or equal to zero ℃ in a future environmental temperature curve and marking the curve as a water icing curve; the water icing curve characteristic acquisition module 5 is used for respectively acquiring the minimum value and the initial time stamp of the water icing curve; the sodium chloride solution concentration obtaining module 6 is used for obtaining the concentration of the sodium chloride solution matched with the minimum value so as to enable the freezing point of the sodium chloride solution to be equal to the minimum value; the preset fault tolerance constant adding module 7 is used for adding a preset fault tolerance constant to the concentration of the sodium chloride solution to obtain the concentration of the sodium chloride solution to be prepared; the sodium chloride solution preparation module 8 is used for completing preparation in the spraying device according to the concentration of the sodium chloride solution to be prepared to obtain a prepared sodium chloride solution; the sodium chloride solution output control module 9 is used for atomizing the prepared sodium chloride solution through the mist outlet based on the starting time stamp and spraying the atomized sodium chloride solution to the road at a preset speed.

Further, the anti-icing device further comprises a target object detection module, a spraying device closing module, a closing time judging module and a manual intervention request generating module which are electrically connected in sequence; the target object detection module is electrically connected with the sodium chloride solution output control module.

The target object detection module is used for respectively detecting whether a target object exists in a preset range of each spraying device or not through a target detection algorithm; the spraying device closing module is used for closing the spraying device with the target object in the preset range if the target object exists; the closing time judging module is used for obtaining the closing time of the spraying device and judging whether the closing time is greater than or equal to a preset time; the manual intervention request generation module is used for generating a manual intervention request and sending the manual intervention request to an external receiving end if the closing time length is greater than or equal to the preset time length.

Further, the target object detection module specifically includes a first target object detection sub-module, a second target object detection sub-module, a third target object detection sub-module, a fourth target object detection sub-module, a fifth target object detection sub-module, a sixth target object detection sub-module, a seventh target object detection sub-module, an eighth target object detection sub-module, a ninth target object detection sub-module, a tenth target object detection sub-module, and an eleventh target object detection sub-module that are electrically connected in sequence; the first target object detection sub-module is electrically connected with the sodium chloride solution output control module, and the eleventh target object detection sub-module is electrically connected with the spraying device closing module.

The first target object detection submodule is used for acquiring image data of a preset range based on the current spraying device; the second target object detection submodule is used for dividing the image data into a plurality of grids through grids with preset density; the third target object detection sub-module is used for respectively predicting a preset number of first-order boundary boxes based on each grid, and each first-order boundary box comprises at least one grid; the fourth target object detection sub-module is used for defining that the target object has the highest confidence; the fifth target object detection submodule is used for respectively acquiring the confidence coefficient of each first-order bounding box; the sixth target object detection submodule is used for acquiring a first-order boundary box with highest confidence coefficient from all confidence coefficients and taking the first-order boundary box as a second-order boundary box; the seventh target object detection submodule is used for calculating the intersection ratio of the second-order boundary frame and the first-order boundary frame respectively; the eighth target object detection submodule is used for reserving a second-order boundary box with the intersection ratio being greater than or equal to a preset ratio as a third-order boundary box; the ninth target object detection submodule is used for reserving a third-order boundary box with highest confidence as a fourth-order boundary box; the tenth target object detection sub-module is used for acquiring the union of all the fourth-order boundary frames, and the union is the detection frame of the target object; the eleventh target object detection submodule is used for judging whether the preset range has a detection frame or not.

Further, the environmental temperature prediction model acquisition module specifically includes a first environmental temperature prediction model acquisition sub-module, a second environmental temperature prediction model acquisition sub-module, a third environmental temperature prediction model acquisition sub-module, a fourth environmental temperature prediction model acquisition sub-module, a fifth environmental temperature prediction model acquisition sub-module, a sixth environmental temperature prediction model acquisition sub-module, a seventh environmental temperature prediction model acquisition sub-module, and an eighth environmental temperature prediction model acquisition sub-module that are electrically connected in sequence; the first environmental temperature prediction model acquisition sub-module is electrically connected with the historical environmental temperature curve acquisition module, and the eighth environmental temperature prediction model acquisition sub-module is electrically connected with the future environmental temperature curve acquisition module.

The first environment temperature prediction model acquisition submodule is used for generating a plurality of acquisition points based on a preset time interval; the second environmental temperature prediction model acquisition submodule is used for respectively acquiring each historical environmental temperature curve based on all acquisition points and generating a data set based on each historical environmental temperature curve; the third environmental temperature prediction model acquisition submodule is used for integrating all data sets to form a total data set; the fourth environmental temperature prediction model acquisition submodule is used for carrying out normalization processing on the total data set to obtain a normalized data set; the fifth environmental temperature prediction model obtaining sub-module is used for dividing the normalized data set into a training set and a verification set according to a preset proportion; the sixth environmental temperature prediction model acquisition submodule is used for defining a topological relation of the neural network model, wherein the topological relation comprises an input layer, an hidden layer and an output layer which are connected in sequence in a signal manner; the seventh environmental temperature prediction model obtaining submodule is used for outputting the training set to the input layer, carrying out iteration for preset times through the neural network model, and respectively obtaining root mean square errors of the verification set and the training result of the current time based on each iteration; the eighth environmental temperature prediction model obtaining sub-module is used for obtaining the minimum value in all root mean square errors, and obtaining a training result corresponding to the minimum value as an environmental temperature prediction model.

Further, the sixth ambient temperature prediction model acquisition submodule is configured to carry a topological relation characterized by the formula (1):

（1）。

wherein,is a neural network model; />Is the->Each input node corresponds to one group of training data of the training set, ++>Is the->Input node to hidden layer +.>Preset weights of the input nodes; />To connect to the +.>A threshold of the input nodes; />Is a transfer function, and。

further, the anti-icing device further comprises a sodium chloride solid mass acquisition module, a solution residual mass acquisition module, a sodium chloride solution concentration acquisition module, a dynamic proportionality coefficient definition module and an optimal value searching module which are electrically connected in sequence; the sodium chloride solid quality acquisition module is electrically connected with the sodium chloride solution output control module.

The sodium chloride solid mass acquisition module is used for acquiring the mass of sodium chloride-containing solid of the sprayed sodium chloride solution; the solution residual mass acquisition module is used for acquiring the evaporation mass of the sprayed sodium chloride solution based on the water ice-forming curve through a fixed integral to obtain the solution residual mass of the sprayed sodium chloride solution; the sodium chloride solution concentration acquisition module is used for calculating the sodium chloride solution concentration on the road through the solid mass of the sodium chloride and the residual mass of the solution; the dynamic proportionality coefficient definition module is used for defining a dynamic proportionality coefficient based on a preset rate and linearly reducing the preset rate through the dynamic proportionality coefficient; the optimal value searching module is used for searching the optimal value of the dynamic proportionality coefficient through a global optimizing algorithm so that the concentration of the sodium chloride solution on the road is equal to the saturation value concentration of the current environment temperature.

Further, the optimal value searching module specifically includes a first optimal value searching sub-module, a second optimal value searching sub-module, a third optimal value searching sub-module, a fourth optimal value searching sub-module, a fifth optimal value searching sub-module and a sixth optimal value searching sub-module which are electrically connected in sequence; the first optimal value searching sub-module is electrically connected with the dynamic proportion coefficient defining module.

The first optimal value searching sub-module is used for giving a plurality of random solutions to the optimal value according to the formula (2), and the result of defining all the random solutions is that the concentration of the sodium chloride solution on the road surface is equal to the saturation value concentration of the current environment temperature.

（2）。

The second optimal value searching sub-module is used for initializing the position of each random solution and respectively updating the current position and the current speed according to the formula (3) based on the same random solution:

（3）。

wherein,is->The random solution is at->Speed of walking->Is->The random solution is at->Speed inertia of steps,/- >Is an inertia coefficient>Is->Self-cognition characterization of the individual random solutions,is->Social cognitive characterization of individual random solutions; />And->Are all the learning factors of the human body,for interval->Random number of->Is->Individual optimal solutions, which have been obtained for the individual random solutions, < >>Globally optimal solution, which has been obtained for all random solutions, < >>Is->Step->And (5) randomly solving.

The third optimal value searching sub-module is used for respectively iterating each random solution for preset times according to the formula (3) and linearly decrementing the inertia coefficient once in each iteration through the formula (4) so as to update each random solutionAnd +.>。

（4）。

The fourth optimal value searching sub-module is used for judging eachWhether the first difference value compared with the previous iteration is smaller than or equal to a first preset adaptation threshold value.

A fifth best value searching submodule for if eachComparing with the first difference value of the previous iteration being smaller than or equal to the first preset adaptation threshold value, respectively judging +.>Whether the second difference value compared with the previous iteration is smaller than or equal to a second preset adaptation threshold value.

The sixth best value searching submodule is used for if And comparing the second difference value of the previous iteration with a second preset adaptation threshold value, and judging that the optimal solution with the optimal value is obtained.

It should be noted that, in this embodiment, the function module embodiment based on the foregoing method embodiment may be referred to the foregoing embodiments for additional content such as optimization, expansion, and illustration of this embodiment, which is not repeated herein.

Fig. 3 is a schematic structural diagram of an electronic device according to an embodiment of the present application. As shown in fig. 3, the electronic device 10 includes a processor 101 and a memory 102 coupled to the processor 101.

The memory 102 stores program instructions for implementing a road anti-icing method of any of the embodiments described above.

The processor 101 is configured to execute program instructions stored in the memory 102 to provide anti-icing of a roadway.

The processor 101 may also be referred to as a CPU (Central Processing Unit ). The processor 101 may be an integrated circuit chip with signal processing capabilities. Processor 101 may also be a general purpose processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.

Further, fig. 4 is a schematic structural diagram of a storage medium according to an embodiment of the present application, and referring to fig. 4, the storage medium 11 according to an embodiment of the present application stores a program instruction 111 capable of implementing all the methods described above, where the program instruction 111 may be stored in the storage medium in the form of a software product, and includes several instructions for making a computer device (which may be a personal computer, a server, or a network device) or a processor (processor) execute all or part of the steps of the methods described in various embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, an optical disk, or other various media capable of storing program codes, or a terminal device such as a computer, a server, a mobile phone, a tablet, or the like.

In the several embodiments provided in the present application, it should be understood that the disclosed systems, apparatuses, and methods may be implemented in other manners. For example, the apparatus embodiments described above are merely illustrative, e.g., the division of elements is merely a logical functional division, and there may be additional divisions of actual implementation, e.g., multiple elements or components may be combined or integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be an indirect coupling or communication connection via some interfaces, devices or units, which may be in electrical, mechanical or other forms.

In addition, each functional unit in each embodiment of the present application may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units. The foregoing is only the embodiments of the present application, and is not intended to limit the scope of the patent application, and all equivalent structures or equivalent processes using the contents of the specification and drawings of the present application or directly or indirectly applied to other related technical fields are included in the scope of the patent protection of the present application.

Claims

1. An anti-icing method for a road, wherein spraying devices for spraying sodium chloride solution with preset concentration are fixedly arranged on one side or two sides of the road, the spraying devices have a solution preparation function, and fog outlets of all the spraying devices face the road, and the anti-icing method is characterized by comprising the following steps:

2. The anti-icing method according to claim 1, characterized in that the prepared sodium chloride solution is atomized through the mist outlet based on the start time stamp and sprayed to the road at a preset rate, after which it comprises:

3. An anti-icing method as claimed in claim 2, wherein detecting the presence of target objects within a predetermined range of each spraying device by means of a target detection algorithm comprises:

acquiring image data of the preset range based on a current spraying device;

defining that the target object has the highest confidence;

judging whether the preset range is provided with the detection frame or not.

4. The method of claim 1, wherein training and learning all historical ambient temperature curves by a machine learning algorithm results in an ambient temperature prediction model comprising:

generating a plurality of acquisition points based on a preset time interval;

Integrating all the data sets to form a total data set;

normalizing the total data set to obtain a normalized data set;

5. The anti-icing method according to claim 4, characterized in that said topological relation is characterized by formula (1):

（1）；

wherein,modeling the neural network; />Is the +.>Each input node, each input nodeA set of training data corresponding to said training set for the entry node,>is the +.>The input node is to the ++>Preset weights of the input nodes; / >For connecting to the +.>A threshold of the input nodes; />Is a transfer function, and->。

6. The anti-icing method according to claim 1, characterized in that the prepared sodium chloride solution is atomized through the mist outlet based on the start time stamp and sprayed to the road at a preset rate, after which it comprises:

7. An anti-icing method according to claim 6, characterized in that finding the optimum value of said dynamic scaling factor by means of a global optimization algorithm so that the concentration of sodium chloride solution of said road surface is equal to the saturation value concentration of the current ambient temperature comprises:

（2）；

wherein,for the set of all random solutions, +.>For each random solution, respectively->Label for random solution->The number of all random solutions; />For the set of velocities of all random solutions, +.>The speed of each random solution;

（3）；

（4）；

Wherein,is->The random solution is at->Inertia coefficient after step optimization, ++>For initial inertia factor, +.>For the current iteration step +.>The maximum iteration step number;

if saidComparing the first difference value of the previous iteration with a first preset adaptation threshold value, and judgingAn optimal solution to the optimal value is determined.

8. An anti-icing device for a road, applied to an anti-icing method for a road according to any of claims 1 to 7, characterized in that it comprises:

9. An electronic device comprising a processor, and a memory coupled to the processor, the memory storing program instructions executable by the processor; the processor, when executing the program instructions stored in the memory, implements the road anti-icing method according to any of claims 1 to 7.

10. A storage medium having stored therein program instructions which when executed by a processor implement a method of anti-icing a road as claimed in any of claims 1 to 7.