Indoor and outdoor seamless positioning method based on IHO

Abstract

The invention discloses an indoor and outdoor seamless positioning method based on an IHO (improved Heima optimization algorithm), and solves the problems of low positioning precision and poor positioning stability of an indoor and outdoor transition area due to the fact that the indoor and outdoor transition area and a complex environment are liable to generate interference on signals and can also shield positioning beacons. According to the method, the optimal weight is calculated through the IHO algorithm according to the WiFi positioning data, the GPS positioning data and the real position data, then the optimal fusion position is calculated based on the optimal weight, accurate positioning of the target to be positioned is obtained, and for indoor and outdoor seamless positioning, the IHO algorithm overcomes the problem that the positioning accuracy is reduced in a transition area by dynamically adjusting the weights of WiFi and GPS.

Description

Technical Field

[0001] This invention relates to a seamless indoor and outdoor positioning method, specifically to a seamless indoor and outdoor positioning method based on IHO. Background Technology

[0002] With the continuous improvement of positioning technology, accurate and continuous location information has become the core foundation for many key applications. Whether it is the management of smart cities, the operation of logistics systems, or the improvement of location-based service experiences, there is an urgent need for positioning technology to go beyond a single environment and achieve high-precision, continuous transitions from outdoor to indoor environments. Therefore, seamless indoor and outdoor positioning has become a research hotspot.

[0003] High-precision positioning in indoor-outdoor transition zones faces several key challenges: When transitioning from outdoors to indoors, the number of visible GNSS satellites decreases, and the geometrical factor of precision (GDOP) increases dramatically, leading to a significant increase in positioning error. Studies show that in transition zones, GNSS positioning errors can rapidly increase from meters to tens of meters, while indoor positioning systems are not yet fully operational, easily creating positioning blind spots. Inertial navigation systems (INS) rely on IMUs (Inertial Measurement Units) to calculate position in transition zones, but errors from gyroscopes and accelerometers accumulate over time, causing positioning drift. In transition zones, UWB or Wi-Fi signals may experience multipath effects or NLOS propagation due to building structures (such as door frames and glass curtain walls), leading to ranging errors. Transition zones are typically densely populated (such as shopping mall entrances and subway stations), making signals susceptible to interference, and moving targets (such as pedestrians and vehicles) may obstruct positioning beacons, affecting stability. Summary of the Invention

[0004] To address the technical problems of insufficient positioning accuracy and poor positioning stability in indoor-outdoor transition areas where complex environments can easily interfere with signals and obstruct positioning beacons, this invention provides an IHO-based seamless indoor-outdoor positioning method.

[0005] To achieve the above objectives, the present invention adopts the following technical solution:

[0006] A seamless indoor and outdoor positioning method based on IHO (Indoor Location and Positioning) is characterized by the following steps:

[0007] Step 1: Collect WiFi and GPS location data of the target to be located, and obtain its real location data according to the preset travel route of the target to be located;

[0008] Step 2: Determine the optimal weights for the fusion positions based on the IHO algorithm;

[0009] 2.1 Initialize the parameters of the IHO algorithm;

[0010] 2.2 Determine the first decision variable using the Piecewise chaotic mapping relationship;

[0011] 2.3 Determine the first fused location based on the first decision variable, WiFi positioning data, and GPS positioning data; calculate the first Euclidean distance between the first fused location and the real location data, and use the first Euclidean distance as the first fitness;

[0012] 2.4. Based on the first fitness, the first decision variable is optimized using the IHO algorithm to obtain the second decision variable;

[0013] 2.5. Determine the second fused location based on the second decision variable, WiFi positioning data, and GPS positioning data; calculate the second Euclidean distance between the second fused location and the real location data, and use the second Euclidean distance as the second fitness.

[0014] 2.6. Compare the second fitness with the first fitness, select the smaller fitness, and determine its corresponding decision variable as the first decision variable;

[0015] 2.7 Repeat steps 2.3-2.6 until the preset number of iterations is reached to obtain the optimal decision variable, and use the optimal decision variable as the optimal weight for the fusion position;

[0016] Step 3: Calculate the optimal fusion position based on the optimal weight to complete the seamless indoor and outdoor positioning of the target to be located.

[0017] Furthermore, step 2.1 specifically includes:

[0018] Initialize the parameters of the IHO algorithm, setting the population size n to 100 and the weight cap X. max The lower bound of the weight is [0 0]. min The parameter dimension dim is 2, the differential evolution scaling factor DE_F is 0.6, and the initial annealing temperature SA_T is 100.

[0019] Furthermore, step 2.2 specifically includes:

[0020] 2.2.1 Generating random numbers z using Piecewise chaotic mapping relationship f+1 f = 1, 2, 3...Max, where Max is the preset number of iterations;

[0021] 2.2.2 Based on random number z f+1 The first decision variable X(:,i) is calculated using the following formula:

[0022] X(:,i)=X min (i)+z f+1 ×(X max (i)-Xmin (i))

[0023] Among them, X min (i), X max (i) represent the lower and upper bounds of the i-th decision variable, respectively.

[0024] Furthermore, step 2.4 specifically includes:

[0025] 2.4.1 Calculate the first candidate solution X_P1(i,:) based on the first decision variable X(:,i):

[0026] X_P1(i,:)=w(t)×X(:,i)+r×(D hippo -IX(:,i))

[0027] Where w(t) is the weight coefficient, t is the current iteration number, r represents a random number in the range of 0 to 1, and D hippo Indicates the hippo position with the best cost in the current iteration; I represents 1 or 2;

[0028] 2.4.2 Calculate the second candidate solution X_P2(i,:) based on the first decision variable X(:,i):

[0029]

[0030] Where v represents a random vector, MG i T represents the average value of randomly selected variables, T represents the probability of selection, and B represents a random number in the range of 1 to 2.

[0031] 2.4.3 Select one of the first candidate solution X_P1(i,:) and the second candidate solution X_P2(i,:) as the first undetermined decision variable X1(:,i) through the simulated annealing selection mechanism;

[0032] 2.4.4 Calculate the third candidate solution X_P3(i,:) based on the first undetermined decision variable X1(:,i):

[0033]

[0034] in, Let represent a random vector with a Levy distribution, and let predator represent the location of the predator or intruder in the search space. denoted by , b represents the distance between the hippopotamus and its predator or intruder; b represents a uniformly distributed random number in the range of 2 to 4; c represents a uniformly distributed random number in the range of 1 to 1.5; d represents a uniformly distributed random number in the range of 2 to 3; g represents a uniformly distributed random number in the range of 2 to 4; u represents a random variable in the range of -1 to 1; and F represents the distance between the hippopotamus and its predator or intruder. predatorLet F be the fitness of the first fitness variable, and let X be the fitness of the threshold variable. min (i)+r×(X max (i)-X min (i)) Generate;

[0035] 2.4.5 Calculate the fourth candidate solution using the differential evolution algorithm;

[0036] 2.4.6. Based on the simulated annealing selection mechanism, select one of the third candidate solution X_P3(i,:) and the fourth candidate solution as the second undetermined decision variable;

[0037] 2.4.7 Calculate the fifth candidate solution X_P5(i,:):

[0038]

[0039] Where h represents a random number that follows a normal distribution;

[0040] 2.4.8. Based on the second undetermined decision variable, the sixth candidate solution is calculated using the elite back-learning algorithm;

[0041] 2.4.9. Select one of the fifth candidate solution X_P5(i,:) and the sixth candidate solution as the second decision variable according to the simulated annealing selection mechanism.

[0042] Furthermore, step 2.2.1 specifically includes:

[0043] Generating random numbers z using Piecewise chaotic mapping relationships f+1 :

[0044]

[0045] Among them, z f It is a chaotic sequence.

[0046] Further, in step 2.4.1, the weighting coefficient w(t) is calculated using the following formula:

[0047]

[0048] Where e is a natural constant, representing exponential operation with e as the base, and q∈[0.2,0.5] is the first preset coefficient.

[0049] Further, in step 2.4.4, the flight parameter β in the Levy distribution is calculated using the following formula:

[0050] β = 1.2 + l × (1 - D)

[0051] in:

[0052] l∈[0.1,0.6] is the second preset coefficient;

[0053]

[0054] X mj Let be the parameter in the m-th row and j-th column of an all-zero matrix X of size n×dim.

[0055] Furthermore, step 3 specifically involves:

[0056] Calculate the optimal fusion position pos based on the optimal weights:

[0057] pos=o1*x1+o2*x2

[0058] Where o1+o2=1, x1 is WiFi positioning data, x2 is GPS positioning data, o1 is the optimal weighting factor on x1, and o2 is the optimal weighting factor on x2;

[0059] Achieve seamless indoor and outdoor positioning of the target.

[0060] Furthermore, in step 2.7:

[0061] The maximum preset number of iterations is in the range of 40-80 times.

[0062] Furthermore, in step 2.4.1, the first preset coefficient q = 0.3;

[0063] In step 2.4.4, the second preset coefficient l = 0.3;

[0064] In step 2.7, the preset number of iterations Max = 50.

[0065] The beneficial effects of this invention are:

[0066] 1. This invention provides an indoor-outdoor seamless positioning method based on IHO (Improved Hippo Optimization Algorithm). It calculates the optimal weights based on WiFi positioning data, GPS positioning data, and real location data, and then calculates the optimal fused position based on the optimal weights to obtain the accurate positioning of the target to be located. For indoor-outdoor seamless positioning, the IHO algorithm overcomes the problem of decreased positioning accuracy in transition areas by dynamically adjusting the weights of WiFi and GPS.

[0067] 2. This invention provides an indoor / outdoor seamless positioning method based on IHO (Indoor-Outdoor Optimization). By integrating Piecewise chaotic mapping, adaptive weighting strategy (step 2.4.1), elite back-learning algorithm (step 2.4.8), and simulated annealing selection mechanism into the IHO algorithm, it improves upon the traditional Hippo Optimization (HO) algorithm. Compared to the traditional HO algorithm, the IHO algorithm can effectively balance global exploration and local exploitation capabilities, overcoming the shortcomings of traditional optimization algorithms that are prone to getting trapped in local optima and have insufficient convergence accuracy.

[0068] 3. This invention provides an indoor / outdoor seamless positioning method based on IHO (Indoor-Outdoor Positioning). It calculates the fused position using decision variables, WiFi positioning data, and GPS positioning data, and determines the fitness by calculating the Euclidean distance with the real location data. Through multiple iterations, it selects the decision variable corresponding to the optimal fitness as a weighting factor. Based on this weighting factor, WiFi positioning data and GPS positioning data are weighted and fused to minimize positioning error. Simulation results show that the IHO algorithm used in this invention has a faster convergence speed and higher accuracy compared to the traditional HO algorithm. Compared to Extended Kalman Filter (EKF), Ant Colony Optimization (ACO), Improved Particle Swarm Optimization (IPSO), and Adaptive Kalman Optimization (AKF), the average positioning error of the IHO algorithm is reduced by 16.48%, 25.04%, 24.56%, and 19.91%, respectively. Attached Figure Description

[0069] Figure 1 This is an experimental scenario diagram from an embodiment of the present invention;

[0070] Figure 2 This is a path comparison diagram of six positioning algorithms in an embodiment of the present invention;

[0071] Figure 3 This is a probability distribution diagram of the cumulative error of the six positioning algorithms in this embodiment of the invention;

[0072] Figure 4 This is a convergence curve diagram of the IHO algorithm and the HO algorithm in the embodiments of the present invention;

[0073] Figure 5 This is a comparison chart of the positioning errors of six positioning algorithms in the embodiments of the present invention. Detailed Implementation

[0074] The technical solution of the present invention will be clearly and completely described below with reference to the accompanying drawings and embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0075] First, it should be noted that the traditional Hippo Optimization (HO) algorithm imitates the collective behavior of hippos in water, their defense mechanisms against predators, and their behavior of escaping predators, transforming them into a mathematical model. In complex optimization problems, it simultaneously achieves global search and local optimization, effectively avoiding getting trapped in local optima.

[0076] This invention provides an indoor / outdoor seamless positioning method based on IHO, comprising the following steps:

[0077] Step 1: Collect WiFi and GPS location data of the target to be located, and obtain its real location data according to the preset travel route of the target to be located.

[0078] Step 2: Determine the optimal weights for the fusion positions based on the IHO algorithm.

[0079] 2.1 Initialize the parameters of the IHO algorithm; the specific settings are as follows:

[0080] Initialize the parameters of the IHO algorithm, setting the population size n to 100 and the weight cap X. max The lower bound of the weight is [0 0]. min The parameter dimension dim is 2, the differential evolution scaling factor DE_F is 0.6, and the initial annealing temperature SA_T is 100.

[0081] 2.2. Determine the first decision variable using the Piecewise chaotic mapping relationship; specifically including:

[0082] 2.2.1 Generating random numbers z using Piecewise chaotic mapping relationship f+1 :

[0083]

[0084] Where f = 1, 2, 3…Max, Max is the preset number of iterations; z f It is a chaotic sequence.

[0085] This embodiment improves the original random number z. f+1 The iterative method enhances the random number z. f+1 Its randomness and diversity provide a wide range of initial solutions, making the distribution of initial decision variables in the algorithm more uniform.

[0086] 2.2.2 Based on random number z f+1 The first decision variable X(:,i) is calculated using the following formula:

[0087] X(:,i)=X min (i)+z f+1 ×(X max(i)-X min (i))

[0088] Among them, X min (i), X max (i) represent the lower and upper bounds of the i-th decision variable, respectively.

[0089] 2.3 Determine the first fused location based on the first decision variable, WiFi positioning data, and GPS positioning data; calculate the first Euclidean distance between the first fused location and the real location data, and use the first Euclidean distance as the first fitness.

[0090] 2.4. Based on the first fitness, the first decision variable is optimized using the IHO algorithm to obtain the second decision variable; specifically including:

[0091] 2.4.1 Calculate the first candidate solution X_P1(i,:) based on the first decision variable X(:,i):

[0092] X_P1(i,:)=w(t)×X(:,i)+r×(D hippo -IX(:,i))

[0093] Where t is the current iteration number, r represents a random number in the range of 0 to 1, and D hippo This indicates the hippo position with the best cost in the current iteration; I represents 1 or 2.

[0094] w(t) is the weighting coefficient, calculated using the following formula:

[0095]

[0096] Where e is a natural constant, representing exponential operation with e as the base, and q∈[0.2,0.5] is the first preset coefficient.

[0097] To enhance the global search capability of the IHO algorithm and prevent it from getting trapped in local optima, in this embodiment, the first preset coefficient q = 0.3. When the weight coefficient is large, the global search capability of the IHO algorithm is strong. As the iteration continues, the weight coefficient gradually decreases, and the IHO algorithm carefully searches at the optimal solution, thus accelerating its convergence speed.

[0098] 2.4.2 Calculate the second candidate solution X_P2(i,:) based on the first decision variable X(:,i):

[0099]

[0100] Where v represents a random vector, MG i T represents the average value of randomly selected variables, and B represents the probability of selection.

[0101] 2.4.3 Select one of the first candidate solution X_P1(i,:) and the second candidate solution X_P2(i,:) as the first undetermined decision variable X1(:,i) through the simulated annealing selection mechanism.

[0102] By simulating the annealing selection mechanism, the IHO algorithm is more likely to accept inferior solutions at high temperatures, thus ensuring population diversity. As the temperature decreases, it gradually tends to accept optimal solutions, promoting convergence and enhancing the algorithm's ability to escape local optima and find the global optimum.

[0103] 2.4.4 Calculate the third candidate solution X_P3(i,:) based on the first undetermined decision variable X1(:,i):

[0104]

[0105] Here, "predator" represents the location of the predator or intruder in the search space. denoted by , b represents the distance between the hippopotamus and its predator or intruder; b represents a uniformly distributed random number in the range of 2 to 4; c represents a uniformly distributed random number in the range of 1 to 1.5; d represents a uniformly distributed random number in the range of 2 to 3; g represents a uniformly distributed random number in the range of 2 to 4; u represents a random variable in the range of -1 to 1; and F represents the distance between the hippopotamus and its predator or intruder. predator Let F be the fitness of the first fitness variable, and let X be the fitness of the threshold variable. min (i)+r×(X max (i)-X min (i))Generate.

[0106] Let represent a random vector with a Levy distribution. The flight parameter β in the Levy distribution is calculated by the following formula:

[0107] β = 1.2 + l × (1 - D)

[0108] in:

[0109] l∈[0.1,0.6] is the second preset coefficient; in this embodiment, the second preset coefficient l=0.3;

[0110]

[0111] X mj Let be the parameter in the m-th row and j-th column of an all-zero matrix X of size n×dim.

[0112] By dynamically adjusting the flight parameter β in the Levy distribution, a large perturbation is provided, giving the IHO algorithm the ability to escape local optima and find optimal candidate solutions.

[0113] 2.4.5 Calculate the fourth candidate solution using the differential evolution algorithm.

[0114] This allows the IHO algorithm to further search for the optimal candidate solution.

[0115] 2.4.6. Based on the simulated annealing selection mechanism, select one of the third candidate solution X_P3(i,:) and the fourth candidate solution as the second undetermined decision variable.

[0116] 2.4.7 Calculate the fifth candidate solution X_P5(i,:):

[0117]

[0118] Where h represents a random number that follows a normal distribution.

[0119] 2.4.8. Based on the second undetermined decision variable, the sixth candidate solution is calculated using the elite back-learning algorithm.

[0120] We use elite reverse learning to perform a fine search and select the optimal candidate solution.

[0121] 2.4.9. Select one of the fifth candidate solution X_P5(i,:) and the sixth candidate solution as the second decision variable according to the simulated annealing selection mechanism.

[0122] 2.5. Determine the second fused location based on the second decision variable, WiFi positioning data, and GPS positioning data; calculate the second Euclidean distance between the second fused location and the real location data, and use the second Euclidean distance as the second fitness.

[0123] 2.6. Compare the second fitness with the first fitness, select the smaller fitness, and determine its corresponding decision variable as the first decision variable.

[0124] 2.7 Repeat steps 2.3-2.6 until the preset number of iterations Max is reached to obtain the optimal decision variable, which is the optimal weight of the fusion position; the preset number of iterations Max ranges from 40 to 80 times; in this embodiment, the preset number of iterations Max = 50.

[0125] Step 3: Calculate the optimal fusion position based on the above optimal weights, specifically as follows:

[0126] Calculate the optimal fusion position pos based on the optimal weights:

[0127] pos=o1*x1+o2*x2

[0128] Where o1+o2=1, x1 is WiFi positioning data, x2 is GPS positioning data, o1 is the optimal weighting factor on x1, and o2 is the optimal weighting factor on x2;

[0129] Achieve seamless indoor and outdoor positioning of the target.

[0130] Experimental Results and Analysis:

[0131] like Figure 1 As shown, the experimental positioning area was a 1-meter square measuring 28 meters long and 8 meters wide. Four wireless access points with strong signals were selected, with coordinates (0.5, 0.5), (2.6, 2), (3, 3.5), and (3.5, 6.5). Data was collected using a "WiFi Magic Box" and a "GPS Test" device. During the experiment, pedestrians moved from indoors to outdoors, with a total of 28 sampling points.

[0132] To verify the positioning performance of the IHO algorithm provided in this embodiment, it is compared with the HO algorithm and four other commonly used positioning algorithms. The positioning trajectory diagrams of each algorithm are shown below. Figure 2 As shown:

[0133] from Figure 2 The comparison of algorithm paths shows that the IHO algorithm is closer to the actual path, indicating that the IHO algorithm is superior to other algorithms in terms of positioning accuracy.

[0134] according to Figure 3 As can be seen, the IHO algorithm has the smallest positioning error and the highest positioning accuracy. The IHO algorithm integrates multiple optimization strategies, achieving a better balance between global search and refined local search, thus outperforming traditional algorithms in terms of positioning error and improving positioning accuracy and robustness.

[0135] according to Figure 4 It can be seen that compared with the HO algorithm, the IHO algorithm has a faster convergence speed. It quickly jumps out of local optima through global search and finds the optimal solution. Through local optimization, the algorithm can accurately locate the optimal solution in the mid-term and find it quickly, thereby improving the convergence speed.

[0136] according to Figure 5 As shown in Table 1, the IHO algorithm has the smallest relative positioning error compared to the other five algorithms.

[0137] Table 1 Comparison of positioning errors under six algorithms

[0138]

[0139] In summary, the IHO algorithm has the lowest average error; its RMSE and standard deviation are relatively low, but close to those of the ACO algorithm and the HO algorithm. In terms of global optimization, the IHO algorithm effectively avoids local optima through simulated annealing and differential evolution, reducing positioning errors. In terms of local fine-tuning, it further reduces errors and improves stability by finely adjusting the solution through elite back-learning.

[0140] In summary, the IHO algorithm exhibits small error fluctuations and strong stability, enabling high-precision continuous positioning in indoor-outdoor transition areas.

[0141] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions within the technical scope disclosed in the present invention should be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A seamless indoor / outdoor positioning method based on IHO, characterized in that, Includes the following steps: Step 1: Collect WiFi and GPS location data of the target to be located, and obtain its real location data according to the preset travel route of the target to be located; Step 2: Determine the optimal weights for the fusion positions based on the IHO algorithm; 2.1 Initialize the parameters of the IHO algorithm; 2.2 Determine the first decision variable using the Piecewise chaotic mapping relationship; 2.3 Determine the first fused location based on the first decision variable, WiFi positioning data, and GPS positioning data; calculate the first Euclidean distance between the first fused location and the real location data, and use the first Euclidean distance as the first fitness; 2.

4. Based on the first fitness, the first decision variable is optimized using the IHO algorithm to obtain the second decision variable; 2.

5. Determine the second fused location based on the second decision variable, WiFi positioning data, and GPS positioning data; calculate the second Euclidean distance between the second fused location and the real location data, and use the second Euclidean distance as the second fitness. 2.

6. Compare the second fitness with the first fitness, select the smaller fitness, and determine its corresponding decision variable as the first decision variable; 2.7 Repeat steps 2.3-2.6 until the preset number of iterations is reached to obtain the optimal decision variable, and use the optimal decision variable as the optimal weight for the fusion position; Step 3: Calculate the optimal fusion position based on the optimal weight to complete the seamless indoor and outdoor positioning of the target to be located.

2. The indoor / outdoor seamless positioning method based on IHO according to claim 1, characterized in that, Step 2.1 specifically involves: Initialize the parameters of the IHO algorithm, setting the population size n to 100 and the weight cap X. max The lower bound of the weight is [0 0]. min [11], parameter dimension dim is 2, differential evolution scaling factor DE_F is 0.6, and annealing initial temperature SA_T is 100.

3. The indoor / outdoor seamless positioning method based on IHO according to claim 2, characterized in that, Step 2.2 specifically includes: 2.2.1 Generating random numbers z using Piecewise chaotic mapping relationship f+1 f = 1, 2, 3...Max, where Max is the preset number of iterations; 2.2.2 Based on random number z f+1 The first decision variable X(:,i) is calculated using the following formula: X(:,i)=X min (i)+z f+1 ×(X max (i)-X min (i)) Among them, X min (i), X max (i) represent the lower and upper bounds of the i-th decision variable, respectively.

4. The indoor / outdoor seamless positioning method based on IHO according to claim 3, characterized in that, Step 2.4 specifically includes: 2.4.1 Calculate the first candidate solution X_P1(i,:) based on the first decision variable X(:,i): X_P1(i,:)=w(t)×X(:,i)+r×(D hippo -IX(:,i)) Where w(t) is the weight coefficient, t is the current iteration number, r represents a random number in the range of 0 to 1, and D hippo Indicates the hippo position with the best cost in the current iteration; I represents 1 or 2; 2.4.2 Calculate the second candidate solution X_P2(i,:) based on the first decision variable X(:,i): Where v represents a random vector, MG i T represents the average value of randomly selected variables, T represents the probability of selection, and B represents a random number in the range of 1 to 2. 2.4.3 Select one of the first candidate solution X_P1(i,:) and the second candidate solution X_P2(i,:) as the first undetermined decision variable X1(:,i) through the simulated annealing selection mechanism; 2.4.4 Calculate the third candidate solution X_P3(i,:) based on the first undetermined decision variable X1(:,i): in, Let represent a random vector with a Levy distribution, and let predator represent the location of the predator or intruder in the search space. denoted by , b represents the distance between the hippopotamus and its predator or intruder; b represents a uniformly distributed random number in the range of 2 to 4; c represents a uniformly distributed random number in the range of 1 to 1.5; d represents a uniformly distributed random number in the range of 2 to 3; g represents a uniformly distributed random number in the range of 2 to 4; u represents a random variable in the range of -1 to 1; and F represents the distance between the hippopotamus and its predator or intruder. predator Let F be the fitness of the first fitness variable, and let X be the fitness of the threshold variable. min (i)+r×(X max (i)-X min (i)) Generate; 2.4.5 Calculate the fourth candidate solution using the differential evolution algorithm; 2.4.

6. Based on the simulated annealing selection mechanism, select one of the third candidate solution X_P3(i,:) and the fourth candidate solution as the second undetermined decision variable; 2.4.7 Calculate the fifth candidate solution X_P5(i,:): Where h represents a random number that follows a normal distribution; 2.4.

8. Based on the second undetermined decision variable, the sixth candidate solution is calculated using the elite back-learning algorithm; 2.4.

9. Select one of the fifth candidate solution X_P5(i,:) and the sixth candidate solution as the second decision variable according to the simulated annealing selection mechanism.

5. The indoor / outdoor seamless positioning method based on IHO according to claim 4, characterized in that, Step 2.2.1 specifically involves: Generating random numbers z using Piecewise chaotic mapping relationships f+1 : Among them, z f It is a chaotic sequence.

6. The indoor / outdoor seamless positioning method based on IHO according to claim 4 or 5, characterized in that, In step 2.4.1, the weighting coefficient w(t) is calculated using the following formula: Where e is a natural constant, representing exponential operation with e as the base, and q∈[0.2,0.5] is the first preset coefficient.

7. The indoor / outdoor seamless positioning method based on IHO according to claim 6, characterized in that, In step 2.4.4, the flight parameter β in the Levy distribution is calculated using the following formula: β＝1.2+l×(1-D) in: l∈[0.1,0.6] is the second preset coefficient; X mj Let be the parameter in the m-th row and j-th column of an all-zero matrix X of size n×dim.

8. The indoor / outdoor seamless positioning method based on IHO according to claim 7, characterized in that, Step 3 specifically involves: Calculate the optimal fusion position pos based on the optimal weights: pos=o1*x1+o2*x2 Where o1+o2=1, x1 is WiFi positioning data, x2 is GPS positioning data, o1 is the optimal weighting factor on x1, and o2 is the optimal weighting factor on x2; Achieve seamless indoor and outdoor positioning of the target.

9. The indoor / outdoor seamless positioning method based on IHO according to claim 8, characterized in that, In step 2.7: The maximum preset number of iterations is in the range of 40-80 times.

10. The indoor and outdoor seamless positioning method based on IHO according to claim 9, characterized in that: In step 2.4.1, the first preset coefficient q = 0.3; In step 2.4.4, the second preset coefficient l = 0.3; In step 2.7, the preset number of iterations Max = 50.

Images