Multi-base-station difference-based Beidou base station site selection optimization method for canyon construction area

Abstract

The invention relates to a canyon building area Beidou base station site selection optimization method based on multi-base station difference, and the method is characterized in that the method comprises the following steps: S1, initializing parameter setting and base station selection; s2, comparing and updating double-difference integer ambiguity; s3, calculating the fitness value of each base station, evaluating the quality of a base station signal according to the fitness value, comparing the fitness value with the fitness value of the passed best position pbest, and if the fitness value is better, updating the pbest; and comparing the fitness values of all the base stations with the fitness value of the global best position gbest, and if the fitness values are better, updating the gbest. Compared with the prior art, through the multi-base-station difference technology and the optimization algorithm, in combination with the double-difference integer ambiguity and the fitness value, optimization of site selection of the Beidou base stations in the canyon construction area is achieved. The positioning precision is improved through double-difference processing, the optimal base station position is dynamically updated through the fitness value and the optimization algorithm, the method is suitable for complex terrains, and base station signal coverage and quality are ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of base station site selection optimization, and particularly to a Beidou base station site selection optimization method for canyon construction areas based on multi-base station differential. Background Art

[0002] The selection of the base station erection location is a key step in the construction site planning and construction. Its importance is reflected in many aspects. The site selection of the base station should fully consider the user needs within the coverage area to ensure that the signal can reach every corner and provide users with stable and high-speed services. A reasonable base station layout can reduce interference, improve network performance, and avoid the increase in interference or excessive measurement errors caused by improper addition of stations. Through scientific site selection, the use efficiency and benefits of the base station can be improved, thereby increasing the return on investment. The selection of the site needs to consider factors such as convenient transportation and good power supply to improve the speed and safety of construction and maintenance. The base station site selection should not only consider the current needs but also the future development, leaving room for future on-site measurement operations and expansion. During the site selection process, environmental protection and energy conservation factors should be considered, and a site with less impact on the environment should be selected, and energy-saving equipment and technologies should be adopted. In the construction of traditional bridge projects, because of the large scale, long cycle, and multiple stages of the project, as an important part of the construction organization design of the project and the beginning of on-site construction, the layout of the general construction plan needs to make reasonable planning and layout of the road traffic, material warehouses, processing sites, main construction machinery, temporary houses, pipelines, etc. at the construction site, so as to correctly handle the spatial relationship between various facilities required during the construction period of the entire construction site and the permanent establishment and the proposed project.

[0003] Bridge construction control networks are usually observed using GPS, and multiple GNSS receivers need to be used for synchronous observation to ensure accuracy. For the construction sites of most bridges, they are located in relatively remote areas where the signals are extremely vulnerable to interference. Before observation, a detailed observation plan needs to be formulated, including the observation period, the measurement accuracy of the antenna height, the data sampling interval, etc., and it should be designed according to factors such as terrain and network size. When detecting the observation quality of the GPS network, high-precision total stations or electro-optical distance measurement side lengths are usually used for combined adjustment to ensure the high precision and uniformity of the control network. There are many places where precision needs to be controlled, and it is very difficult to effectively ensure that the signals of GNSS receivers are not interfered. Moreover, currently, more and more positioning methods are based on RTK devices for positioning, and most commercial RTKs use network RTK positioning technology. The connected base stations are usually more than 20 kilometers away. The signals of the receivers are very easily affected by multipath errors, the ionosphere, and the troposphere. In this regard, we need to set up several static base stations at the construction site to receive satellite signals and transmit them to the receivers in real time, so as to ensure that the signals are not interfered. When selecting the layout control of static base stations, since the quality of satellite data received at each point is different, we need to ensure that the layout control is optimized after excluding unreasonable points, so as to achieve the best layout control and ensure the integrity of data quality to the greatest extent. It is necessary to optimize the selection of the layout positions of base stations at the bridge construction site.

[0004] Regarding the optimal siting of base stations, Patent CN118488465A presents a method for optimizing the siting of base stations based on a Transformer network. Patent CN110972152A provides a method for optimizing the siting of 5G base stations in urban spaces considering signal blocking effects. CN112929891A provides a method for simulating and optimizing the siting of 5G base station signals based on an indoor three-dimensional structured model, with relatively low control precision and poor anti-interference ability. Regarding the particle swarm algorithm, Patent CN202010770387.4 provides a method for optimizing the siting of substations based on a hybrid algorithm of barycentric regression and particle swarm. First, determine the number n of new substations to be built in the power grid, then use the barycentric regression algorithm to divide the load nodes in the power grid into n load areas, obtain the position coordinates of the substations supplying power to each load area and the loads they supply, and use them as the initial values of the particle swarm algorithm to initialize the positions of the particles. Then, with the minimum global load moment as the target fitness value, use the particle swarm algorithm to optimize the positions of the substations and the loads they supply, obtain the optimized positions of the n substations and the loads they supply, and finally obtain the transformation capacities of each substation according to the constraint relationship between the transformation capacity of the substation and the load it supplies. However, its positioning accuracy is low and its anti-interference ability is weak. Summary of the Invention

[0005] The purpose of the present invention is to overcome the above-mentioned defects existing in the prior art, and provide a Beidou base station site selection optimization method for canyon construction areas based on multi-base station differential. Through multi-base station differential technology and optimization algorithms, combined with double-difference integer ambiguity and fitness values, the optimization of Beidou base station site selection in canyon construction areas is realized. By double-difference processing, the positioning accuracy is improved, and the optimal base station position is dynamically updated by using the fitness value and optimization algorithm, which is applicable to complex terrains and ensures the base station signal coverage and quality.

[0006] The present invention optimally selects the base station layout positions at the bridge construction site through a multi-base station double-difference integer ambiguity anomaly recognition method, and optimally selects the base station layout positions at the construction site through the random walk algorithm in the particle swarm algorithm.

[0007] The particle swarm algorithm is an optimization calculation method based on swarm intelligence. Since its proposal, the particle swarm algorithm has been widely used in the fields of function optimization, neural network training, fuzzy system control, etc. due to its advantages such as simple implementation and fast convergence speed. By selecting two base stations with better signals as standards, randomly selecting points in the layout area as initial base stations, calculating the fitness value of each base station, evaluating the quality of the base station signals according to the fitness value, for each base station, comparing its fitness value with the fitness value of the best position pbest it has experienced, if it is better, then update pbest; at the same time, comparing the fitness values of all base stations with the fitness value of the global best position gbest, if it is better, then update gbest. The present invention optimizes the bridge construction layout based on the multi-base station double-difference integer ambiguity particle swarm algorithm.

[0008] The selection of the standard base station positions should be that the surrounding of the base station without obstacles should be open, the cut-off elevation angle should exceed 15 degrees, and there should be no signal reflectors around (such as large water areas, large buildings, etc.) to reduce problems such as multipath interference. It is best that there is no large obstacle (such as high-rise buildings, mountains, etc.) between the base station radio antenna and the rover antenna to avoid affecting the signal transmission distance. The rooftop station is widely used because it is easier to meet the construction requirements. Therefore, the rooftop station can be selected as the position of the standard base station. After setting up the rooftop base station, they can observe the satellites, calculate the double-difference integer ambiguity between them, and the ratio (threshold) corresponding to the fixed double-difference integer ambiguity. When the ratio value is greater than 3, it is an effective fixed solution, and the larger the ratio value, the more reliable the fixed solution. Therefore, the ratio value can be used as a fitness value for reference; in addition, the double-difference integer ambiguity values in the same area will not be too large, and the absolute value of the double-difference ambiguity between two base stations should be as small as possible. Therefore, the double-difference ambiguity can also be considered as a fitness value.

[0009] The present invention is oriented towards precise positioning before civil engineering construction. By combining the particle swarm algorithm and double-difference integer ambiguity, the point with the most stable signal and the least GNSS signal transmission interference in an area is selected as the control point of the static base station. After obtaining the control points of these several static base stations, base stations at a relatively long distance are no longer used as control points. Relevant positioning can be carried out through some self-made small positioning software. The optimized selected points are used as the base stations of the small positioning software for differencing, and accurate positioning accuracy can still be obtained. The self-made small positioning software can be placed on the safety helmet of the worker to check whether the worker constructs in a standard manner, or placed on the hoisted bridge beam segment to realize the pose change of the beam segment during the hoisting process.

[0010] The present invention provides a Beidou base station site selection optimization method based on multi-base station differencing for canyon construction areas, including the following steps:

[0011] S1: Initial parameter setting and base station selection: Randomly select n points in a pre-construction site, set initial parameters for each point, and establish the initial double-difference integer ambiguity through the observation equation; select the standard base station; establish the initial double-difference integer ambiguity through the observation equation. Establishing the observation equation means establishing the carrier phase observation equation. For the carrier phase observation values of the same satellite at two different observation stations, list the corresponding observation equations, which include the double-difference term; select the standard base station;

[0012] S2: Comparison and update of double-difference integer ambiguity: Through multi-base station differencing, that is, comparing and differencing the observation data of multiple base stations, compare the initial double-difference integer ambiguity with the double-difference integer ambiguity of the standard base station. Based on the least squares method, obtain the fixed solution of the initial double-difference integer ambiguity. If the fixed solution of the initial double-difference integer ambiguity is better than the double-difference integer ambiguity of the standard base station, then the point where the fixed solution of the double-difference integer ambiguity is better than the double-difference integer ambiguity of the standard base station is used as the standard base station point; compare the double-difference integer ambiguity of the point with a better fixed solution with the double-difference integer ambiguity of the best position passed by the point with a better fixed solution. If it is better, then the point with a better fixed solution is used as the current best position;

[0013] S3: Fitness value calculation and optimization, find the optimal solution through iteration: Calculate the fitness value of each base station based on the double-difference integer ambiguity, evaluate the quality of the base station signal according to the fitness value, and based on multi-base station differencing, comprehensively consider each base station and find the optimal solution through continuous iteration. For each base station, compare its fitness value with the fitness value of the best position pbest it has experienced. If it is better, then update pbest; at the same time, compare the fitness values of all base stations with the fitness value of the global best position gbest. If it is better, then update gbest.

[0014] Calculate the fitness value of each base station, and compare and update it with the fitness value of the best position pbest it has experienced and the fitness value of the global best position gbest, that is, based on multi-base station difference, comprehensively consider the situation of each base station and the mutual relationship between them, and find the optimal solution through continuous iteration.

[0015] The best position pbest that each base station has experienced refers to the position with the optimal fitness value (that is, representing the best base station signal, the best positioning effect, etc.) among all the positions the base station has reached before. For each base station, during the entire iterative optimization process, it will continuously move to different positions (that is, different points randomly selected in the pre-construction site), and calculate the corresponding fitness value at each position.

[0016] Further, in S1, select standard base stations: Select two base stations with high signal strength as standard base stations.

[0017] Further, in S1, the initialization parameters include: coordinate parameters, observation data parameters, satellite orbit parameters. The initialization parameters include: coordinate parameters, observation data parameters, satellite orbit parameters; the coordinate parameters include the initial values of the geodetic coordinates (longitude, latitude, altitude, etc.) of the base station; the observation data parameters include the initial settings of pseudo-range observations, carrier phase observations, etc. of satellite signals.

[0018] Further, in S3, in each iteration, the base station updates itself by tracking two extreme values pbest and gbest.

[0019] Further, in S3, after finding the two optimal values pbest and gbest, based on the formula for updating the velocity of PSO particles, the base station updates its velocity and position through the following formula:

[0020] v i = ω × v i + c1 × rand() × (pbest i - x i ) + c2 × rand() × (gbest i - x i );

[0021] x i = x i + v i

[0022] Where ω is the inertia factor, and its value is non - negative. When its value is larger, the global optimization ability is strong and the local optimization ability is weak. When its value is smaller, the global optimization ability is weak and the local optimization ability is strong. Dynamic ω can obtain better optimization results than the fixed value. Dynamic ω changes linearly during the PSO search process or changes dynamically according to a certain measure function of the PSO performance. The measure function is the fitness variance, average fitness or optimal fitness change rate.

[0023] Where i = 1, 2, …, N, and N is the total number of base stations in this group.

[0024] v i is the velocity of the base station, and the maximum value of v i is V max > 0. If v i > V max , then v i = V max .

[0025] rand() is a random number between (0, 1).

[0026] c1 and c2 are learning factors.

[0027] x i is the current position of the base station.

[0028] pbest i is the best position the base station has experienced; gbest i is the global best position of the base station.

[0029] Furthermore, in S2, the fixed solution of the initial double - difference integer ambiguity is obtained based on the least - squares method. Specifically: The least - squares method determines the unknown parameters by minimizing the sum of the squares of the errors between the observed values and the theoretical values, which is consistent with the requirements for solving the double - difference integer ambiguity. An observation equation containing the double - difference integer ambiguity is established according to the double - difference observations for parameter estimation by the least - squares method. A least - squares objective function is constructed, and the least - squares estimated value of the double - difference integer ambiguity is obtained. The obtained least - squares estimated value of the double - difference integer ambiguity is tested and screened to determine the final fixed solution of the initial double - difference integer ambiguity.

[0030] Furthermore, in S2, the method for testing and screening the obtained least - squares estimated value of the double - difference integer ambiguity is as follows: Combining the integer property of the ambiguity, the rounded - off operation is performed on the least - squares estimated value of the ambiguity to obtain the candidate value of the integer ambiguity. Then, the change in the objective function value before and after rounding is calculated. If the change is small, it is determined as the final fixed solution of the initial double - difference integer ambiguity.

[0031] Further, in S3, if there are multiple base stations with better fitness values, select the base station with the largest fitness value, that is, among the multiple base stations with fitness values better than gbest, select the position of the base station with the largest fitness value as the new global best position gbest.

[0032] Further, in the formula for updating the PSO particle velocity, the inertia weight ω(t) is adjusted by the linearly decreasing weight strategy, and the adjustment formula for the inertia weight ω(t) is:

[0033] ω (t) =(ω int -ω end )(G k -g) / G k +ω end ;

[0034] where G k is the maximum number of iterations; ω ini is the initial inertia weight;

[0035] ω end is the inertia weight when iterating to the maximum number of evolutionary generations.

[0036] Further, in S3, the iteration termination condition is selected as the maximum number of iterations Gk or / and the optimal position searched by the base station group so far satisfies the predetermined minimum fitness threshold.

[0037] Compared with the prior art, the present invention has the following advantages:

[0038] (1) Through the multi-base station differential technology and optimization algorithm, combined with the double-difference integer ambiguity and fitness value, the present invention realizes the optimization of the Beidou base station site selection in the canyon construction area. The positioning accuracy is improved through double-difference processing, and the optimal base station position is dynamically updated using the fitness value and optimization algorithm, which is applicable to complex terrains and ensures the base station signal coverage and quality.

[0039] (2) Aiming at the complex environment of the canyon terrain, the base station site selection is optimized to ensure signal coverage and quality; combined with the high-precision positioning ability of the Beidou system, the precise optimization of the base station site selection is realized; it is used for high-precision positioning and monitoring in engineering construction to improve construction efficiency and quality. The multi-base station differential can adapt to the complex signal propagation in the canyon. Based on the least squares method and ambiguity comparison, the canyon ambiguity problem can be overcome, and the fitness value calculation and iterative optimization can comprehensively evaluate the complex environmental factors in the canyon.

[0040] (3) Double-difference integer ambiguity: By performing double-difference processing, errors such as receiver clock error, satellite clock error, and atmospheric delay are eliminated. The integer value of the integer ambiguity is determined using the ambiguity fixing technique to improve positioning accuracy. The fitness value can be defined as the sum of the squares of the residuals of the double-difference integer ambiguity. The smaller the residual, the higher the fitness value. Through the update mechanism of pbest and gbest, the base station location is gradually optimized. Combining random point distribution and optimized screening ensures the scientificity and practicality of the base station site selection. Description of the Drawings

[0041] Figure 1 It is a flowchart of a Beidou base station site selection optimization method for canyon construction areas based on multi-base station differential. Specific Embodiments

[0042] The present invention will be described in detail below with reference to the drawings and specific embodiments. Features such as component models, material names, connection structures, control methods, algorithms, etc. that are not clearly described in this technical solution are regarded as common technical features disclosed in the prior art.

[0043] Embodiment 1

[0044] This embodiment provides a Beidou base station site selection optimization method for canyon construction areas based on multi-base station differential, including the following steps:

[0045] S1: Initial parameter setting and base station selection: Randomly select n points in a pre-construction site, perform initial parameter setting for each point, and obtain the initial double-difference integer ambiguity through the establishment of an observation equation; select standard base stations; obtain the initial double-difference integer ambiguity through the establishment of an observation equation. The establishment of the observation equation means establishing a carrier phase observation equation. For the carrier phase observation values of the same satellite at two different observation stations, the corresponding observation equation is listed, which includes double-difference terms; select two base stations with high signal strength as standard base stations.

[0046] The initial parameters include: coordinate parameters, observation data parameters, and satellite orbit parameters. The initial parameters include: coordinate parameters, observation data parameters, and satellite orbit parameters; the coordinate parameters include initial values such as the geodetic coordinates (longitude, latitude, altitude) of the base station; the observation data parameters include the initial settings of pseudo-range observation values and carrier phase observation values of satellite signals.

[0047] S2: Double-difference integer ambiguity comparison and update: Through multi-base station differencing, that is, comparing and differencing the observation data of multiple base stations, comparing the initial double-difference integer ambiguity with the double-difference integer ambiguity of the standard base station, obtaining the fixed solution of the initial double-difference integer ambiguity based on the least squares method. If the fixed solution of the initial double-difference integer ambiguity is better than the double-difference integer ambiguity of the standard base station, then the point where the fixed solution of the double-difference integer ambiguity is better than that of the standard base station is used as the standard base station point; when comparing the double-difference integer ambiguity of the point with a better fixed solution with the double-difference integer ambiguity of the best position passed by the point with a better fixed solution, if it is better, then the point with a better fixed solution is used as the current best position;

[0048] Based on the least squares method, the fixed solution of the initial double-difference integer ambiguity is obtained. Specifically: The least squares method determines the unknown parameters by minimizing the sum of the squares of the errors between the observed values and the theoretical values, which is consistent with the requirements for solving the double-difference integer ambiguity; an observation equation containing the double-difference integer ambiguity is established according to the double-difference observations for parameter estimation by the least squares method; a least squares objective function is constructed, and the least squares estimated value of the double-difference integer ambiguity is obtained, and the obtained least squares estimated value of the double-difference integer ambiguity is tested and screened to determine the final fixed solution of the initial double-difference integer ambiguity.

[0049] The method for testing and screening the obtained least squares estimated value of the double-difference integer ambiguity is to combine the integer characteristics of the ambiguity, perform rounding operations on the ambiguity estimated value obtained by the least squares method to obtain candidate integer ambiguity values, and then calculate the change in the objective function value before and after rounding. If the change is small, it is determined as the final fixed solution of the initial double-difference integer ambiguity.

[0050] S3: Fitness value calculation and optimization, finding the optimal solution through iteration: Calculate the fitness value of each base station based on the double-difference integer ambiguity, evaluate the quality of the base station signal according to the fitness value, and based on multi-base station differencing, comprehensively consider each base station and find the optimal solution through continuous iteration. For each base station, compare its fitness value with the fitness value of the best position pbest it has experienced. If it is better, then update pbest; at the same time, compare the fitness values of all base stations with the fitness value of the global best position gbest. If it is better, then update gbest.

[0051] Calculate the fitness value of each base station, and compare and update it with the fitness value of the best position pbest it has experienced and the fitness value of the global best position gbest, that is, based on multi-base station differencing, comprehensively consider the situations of each base station and their mutual relationships, and find the optimal solution through continuous iteration.

[0052] The best position pbest experienced by each base station refers to the position with the optimal fitness value (i.e., representing the best base station signal and the best positioning effect, etc.) among all the positions the base station has reached before. For each base station, during the entire iterative optimization process, it continuously moves to different positions (i.e., different points randomly selected in the pre-construction site), and calculates the corresponding fitness value at each position.

[0053] In the specific implementation, after finding the two optimal values of pbest and gbest, based on the formula for updating the PSO particle velocity, the base station updates its velocity and position through the following formula:

[0054] v i = ω × v i + c1 × rand() × (pbest i - x i ) + c2 × rand() × (gbest i - x i );

[0055] x i = x i + v i

[0056] Where ω is the inertia factor, and its value is non - negative; when its value is larger, the global optimization ability is strong and the local optimization ability is weak; when its value is smaller, the global optimization ability is weak and the local optimization ability is strong; dynamic ω can obtain better optimization results than a fixed value; dynamic ω changes linearly during the PSO search process or changes dynamically according to a certain measure function of the PSO performance, and the measure function is the fitness variance, average fitness or optimal fitness change rate;

[0057] Where i = 1, 2,..., N, and N is the total number of base stations in this group;

[0058] v i is the velocity of the base station, and the maximum value of v i is V max > 0. If v i > V max then v i = V max ;

[0059] rand() is a random number between (0, 1);

[0060] c1 and c2 are learning factors, usually c1 = c2 = 2;

[0061] x i is the current position of the base station;

[0062] pbesti is the best location experienced by the base station; gbest i is the global best location of the base station.

[0063] v in the formula i represents the first part, called the memory term, indicating the influence of the previous velocity magnitude and direction;

[0064] c1×rand()×(pbest in the formula j -x j ) is called the self-cognition term, which is a vector pointing from the current point to the particle's own best point, indicating the part of the particle's action that comes from its own experience;

[0065] c2×rand()×(gbest in the formula j -x j ) is called the swarm cognition term, which is a vector pointing from the current point to the best point of the population, reflecting the cooperation and knowledge sharing among particles. The particle decides its next movement based on its own experience and the best experience among its peers.

[0066] If there are multiple base stations with better fitness values, select the base station with the largest fitness value, that is, among multiple base stations with fitness values better than gbest, select the location of the base station with the largest fitness value as the new global best location gbest.

[0067] In the formula for updating the PSO particle velocity, the inertia weight ω is adjusted through the linearly decreasing weight strategy (t) and the adjustment formula for the inertia weight ω (t) is as follows:

[0068] ω (t) =(ω ini -ω end )(G k -g) / G k +ω end ;

[0069] where G k is the maximum number of iterations; ω ini is the initial inertia weight;

[0070] ω end is the inertia weight at the maximum number of evolutionary generations.

[0071] The typical weights are: ω ini =0.9, ω end =0.4.

[0072] (i) The introduction has greatly improved the performance of the PSO algorithm, and the global and local search capabilities can be adjusted for different search problems.

[0073] The iteration termination condition is selected as the maximum number of iterations Gk or / and the optimal position found by the base station group so far satisfies a predetermined minimum adaptation threshold. Components not elaborated in this embodiment are all existing components that can be purchased through public channels.

[0074] The above description of the embodiments is to enable those of ordinary skill in the art to understand and use the invention. It is obvious that those skilled in the art can easily make various modifications to these embodiments and apply the general principles described herein to other embodiments without creative efforts. Therefore, the present invention is not limited to the above embodiments, and the improvements and modifications made by those skilled in the art without departing from the scope of the present invention should be within the protection scope of the present invention.

Claims

1. A Beidou base station site selection optimization method for canyon construction areas based on multi-base station differential, characterized in that Including the following steps: S1: Initial parameter setting and base station selection: Randomly select n points in a pre-construction site, perform initial parameter setting for each point, and establish the initial double-difference integer ambiguity through the observation equation; Select the standard base station; S2: Double-difference integer ambiguity comparison and update: Through multi-base station differencing, compare the initial double-difference integer ambiguity with the double-difference integer ambiguity of the standard base station, and obtain the fixed solution of the initial double-difference integer ambiguity based on the least squares method; If the fixed solution of the initial double-difference integer ambiguity is better than the fixed solution of the double-difference integer ambiguity of the standard reference base station, take the point with the better fixed solution as the new standard reference base station, and then compare the double-difference integer ambiguity of the point with the better fixed solution with the ambiguity of the best position passed by the point with the better fixed solution. If it is better, update the current best position; S3: Fitness value calculation and optimization, find the optimal solution through iteration: Calculate the fitness value of each base station based on the double-difference integer ambiguity, evaluate the quality of the base station signal according to the fitness value, and based on multi-base station differencing, for each base station, compare its fitness value with the fitness value of the best position pbest it has experienced. If it is better, update pbest; At the same time, compare the fitness values of all base stations with the fitness value of the global best position gbest. If it is better, update gbest.

2. The method for optimizing the siting of Beidou base stations in canyon construction areas based on multi-base station differential according to claim 1, wherein, In S1, select the standard base station: Select two base stations with high signal strength as the standard base stations.

3. The Beidou base station site selection optimization method for canyon construction areas based on multi-base station differential according to claim 1, characterized in that In S1, the initial parameters include: coordinate parameters, observation data parameters, satellite orbit parameters.

4. A Beidou base station site selection optimization method for canyon construction areas based on multi-base station differential, characterized in that, In S3, in each iteration, the base station updates itself by tracking two extreme values pbest and gbest.

5. The Beidou base station site selection optimization method for canyon construction areas based on multi-base station differential according to claim 1, characterized in that, In S3, after finding the two optimal values of pbest and gbest, based on the formula for updating the PSO particle velocity, the base station updates its velocity and position through the following formula, and the formula is: v i = ω × v i + c1 × rand() × (pbest i - x i ) + c2 × rand() × (gbest i - x i ); x i = x i + v i Where ω is the inertia factor; The dynamic ω changes linearly during the PSO search process or changes dynamically according to a certain measure function of the PSO performance. The measure function is the fitness variance, average fitness, or optimal fitness change rate; Where i = 1, 2,..., N, and N is the total number of base stations in this group; v i is the speed of the base station, v i The maximum value of which is V max > 0, if v i >V max then v i =V max ; rand() is a random number between (0, 1); c1 and c2 are learning factors; x i is the current location of the base station; pbest i The best location experienced by the base station; gbest i The global best location of the base station.

6. The Beidou base station site selection optimization method for canyon construction areas based on multi-base station differential according to claim 1, characterized in that In S2, obtain the fixed solution of the initial double-difference integer ambiguity based on the least squares method. Specifically: The least squares method determines the unknown parameters by minimizing the sum of the squares of the errors between the observed values and the theoretical values, which is consistent with the requirement for solving the double-difference integer ambiguity; Establish an observation equation containing the double-difference integer ambiguity based on the double-difference observations for parameter estimation by the least squares method; Construct a least squares objective function, solve to obtain the least squares estimated value of the double-difference integer ambiguity, and perform inspection and screening on the obtained double-difference integer ambiguity estimated value to determine the final fixed solution of the initial double-difference integer ambiguity.

7. A Beidou base station site selection optimization method for canyon construction areas based on multi-base station differential, characterized in that In S2, the method for testing and screening the obtained double-difference integer ambiguity estimation value is as follows: combining the integer characteristics of the ambiguity, performing rounding operation on the ambiguity estimation value obtained by the least squares method to obtain integer ambiguity candidate values, and then calculating the change in the objective function value before and after rounding, and then determining it as the final initial double-difference integer ambiguity fixed solution.

8. The Beidou base station site selection optimization method for canyon construction areas based on multi-base station differential according to claim 1, characterized in that In S3, if there are multiple base stations with better fitness values, select the base station with the largest fitness value, that is, among multiple base stations with fitness values better than gbest, select the position of the base station with the largest fitness value as the new global best position gbest.

9. The optimization method for Beidou base station site selection in canyon construction areas based on multi-base station differential according to claim 5, characterized in that In the formula for updating the velocity of PSO particles, the inertia weight ω is adjusted by the linearly decreasing weight strategy (t) for adjustment. The adjustment formula for the inertia weight ω (t) is as follows: ω (t) = (ω ini - ω end )(G k - g) / G k + ω end ; Among them, G k is the maximum number of iterations; ω ini is the initial inertia weight; ω end is the inertia weight value when iterating to the maximum number of evolutionary generations.

10. A method for optimizing the siting of Beidou base stations in canyon construction areas based on multi-base station differential, characterized in that, In S3, the iteration termination condition is selected as the maximum number of iterations Gk or / and the optimal position searched by the base station group so far satisfies the predetermined minimum fitness threshold.

Images