CN110493731B

CN110493731B - Movement track obtaining method and device, storage medium and equipment

Info

Publication number: CN110493731B
Application number: CN201910848820.9A
Authority: CN
Inventors: 陈毅臻; 吴汉杰; 宋翔宇; 鲁梦平; 师婷婷; 戴云峰
Original assignee: Tencent Technology Shenzhen Co Ltd
Current assignee: Tencent Technology Shenzhen Co Ltd
Priority date: 2019-09-09
Filing date: 2019-09-09
Publication date: 2022-03-29
Anticipated expiration: 2039-09-09
Also published as: CN110493731A

Abstract

The application discloses a moving track obtaining method, a moving track obtaining device, a storage medium and equipment, and belongs to the field of wireless network communication. The method comprises the following steps: sampling a target geographic area to obtain a plurality of samples representing possible positions of a mobile object in the target geographic area; at each moment of position estimation of the moving object after sampling is finished, acquiring the position information of the generated sample at the current moment based on the position information of the generated sample at the previous moment; acquiring the signal intensity matched with the identity of the terminal carried by the mobile object at the current moment, and acquiring the weight of the generated sample at the current moment based on the acquired signal intensity and the path loss model; determining the appearance position of the mobile object in the target geographic area at the current moment based on the generated position information and the weight of the sample at the current moment; and connecting the appearance positions of the moving objects in the target geographic area at various moments according to the time sequence to obtain a moving track. Fast multiplexing and large scale expansion are supported.

Description

Movement track obtaining method and device, storage medium and equipment

Technical Field

The present application relates to the field of wireless network communication technologies, and in particular, to a method, an apparatus, a storage medium, and a device for obtaining a movement trajectory.

Background

With the rapid development of wireless networks and the rapid popularization of mobile terminals, indoor positioning technologies based on wireless networks and mobile terminals play a great role in the business field. For example, the moving track of the pedestrian in the mall can be obtained through an open wireless network in the mall and a mobile terminal held by the pedestrian in the mall, and then based on the obtained moving track, operations such as analyzing user interests and hobbies and optimizing user experience are executed.

Continuing with the mall as an example, the process of acquiring the moving track of the pedestrian in the related art is generally as follows: the method comprises the steps that a professional holds professional equipment in a hand in advance to collect RSSI (Received Signal Strength Indication) values at each reference position point in a shopping mall, and a set of the collected RSSI values is collected into a database as fingerprint information. In the positioning stage, after fingerprint information of a pedestrian passing the mobile terminal test is obtained, the fingerprint information is compared with fingerprint information stored in a database based on similarity, the position where the most similar fingerprint information is located is determined as the position of the pedestrian, and then the moving track of the pedestrian is obtained based on the positioned position of the pedestrian.

For the above-mentioned moving track obtaining method, because fingerprint information needs to be collected in advance at each reference position point in the mall, the offline workload is large, and it is difficult to reuse and expand on a large scale, i.e. the universality is poor, and in addition, when the arrangement in the mall is changed, fingerprint information needs to be collected again, and the effect is poor.

Disclosure of Invention

The embodiment of the application provides a movement track obtaining method, a movement track obtaining device, a storage medium and a device, and solves the problems that in the related art, fingerprint information collection needs to be carried out on each reference position point in a shop floor in advance, and therefore offline workload is large, multiplexing and large-scale expansion are difficult, and the effect is poor. The technical scheme is as follows:

in one aspect, a method for acquiring a movement trajectory is provided, where the method includes:

sampling a target geographic area to obtain a plurality of samples representing possible positions of a mobile object in the target geographic area;

at each moment of position estimation of the moving object after sampling is finished, acquiring the position information of the generated sample at the current moment based on the position information of the generated sample at the previous moment;

acquiring the signal intensity matched with the identity of the terminal carried by the mobile object at the current moment, and acquiring the weight of the generated sample at the current moment based on the acquired signal intensity and the path loss model;

determining the appearance position of the mobile object in the target geographic area at the current moment based on the generated position information and the weight of the sample at the current moment;

and connecting the appearance positions of the mobile objects in the target geographic area at various moments according to the time sequence to obtain the moving track of the mobile objects.

In another aspect, a movement trajectory acquisition apparatus is provided, the apparatus including:

the system comprises a sample initialization module, a position estimation module and a position estimation module, wherein the sample initialization module is used for sampling a target geographic area to obtain a plurality of samples representing possible positions of a mobile object in the target geographic area;

the sample position updating module is used for acquiring the position information of the generated sample at the current moment based on the position information of the generated sample at the previous moment at each moment of position estimation of the moving object after sampling is finished;

the sample weight updating module is used for acquiring the signal strength matched with the identity of the terminal carried by the mobile object at the current moment at each moment of position estimation of the mobile object after sampling is finished, and acquiring the weight of the generated sample at the current moment based on the acquired signal strength and the path loss model;

the position estimation module is used for determining the appearance position of the mobile object in the target geographic area at the current moment based on the position information and the weight of the generated sample at the current moment at each moment of position estimation of the mobile object after sampling is finished;

and the moving track generating module is used for connecting the appearance positions of the moving objects in the target geographic area at all times according to the time sequence to obtain the moving track of the moving objects.

In a possible implementation manner, the sample weight updating module is further configured to obtain raw data detected by each access point deployed in the target geographic area, where the raw data includes an identity of a detected terminal and a signal strength at a position of the detected terminal, and one terminal corresponds to one moving object; and acquiring the signal intensity matched with the identity of the terminal carried by the mobile object at the current moment from the detected original data.

In one possible implementation, the apparatus further includes:

the data preprocessing module is used for preprocessing the detected original data;

the sample weight updating module is further configured to obtain, in the preprocessed original data, a signal strength at a current time that matches an identity of a terminal carried by the mobile object;

wherein the raw data further comprises a timestamp, and the data preprocessing comprises at least one of:

filtering redundant data in the original data, wherein the redundant data at least comprises abnormal data and data irrelevant to the moving object;

and dividing the signal intensity data included in the original data according to the identity identifier and the timestamp of the detected terminal to form a corresponding relation with the identity identifier and the timestamp as indexes and the signal intensity data as values.

In a possible implementation manner, the sample weight updating module is further configured to, for each sample, obtain a weight of the sample at a previous time; acquiring the signal strength detected by each access point at the position of the sample from the acquired signal strength; obtaining likelihood probabilities of signal strengths detected at locations where the samples are based on the path loss model; acquiring the weight of the sample at the current moment based on the weight of the sample at the previous moment and the likelihood probability;

wherein the weight of the sample at the current time instant is proportional to the likelihood probability.

In one possible implementation, the apparatus further includes;

the parameter estimation module is used for acquiring the values of the reference signal strength and the path loss index in the path loss model;

the parameter estimation module is further configured to, when a distance between access points deployed in the target geographic area is smaller than a distance threshold, obtain signal strength data collected when the access points perform mutual detection, and fit values of the reference signal strength and the path loss index according to the distance between the access points and the collected signal strength data; or when the distance between the access points is greater than the distance threshold, acquiring signal intensity data acquired by the detection equipment when the detection equipment detects the access points at different positions; and acquiring values of the reference signal strength and the path loss exponent based on the signal strength data acquired at different positions.

In a possible implementation manner, the sample location updating module is further configured to obtain a state transition model; and acquiring the position information of the generated sample at the current moment based on the position information of the generated sample at the previous moment and the state transition model.

In a possible implementation manner, the sample position updating module is further configured to obtain a time interval between a previous time and a current time, and obtain a maximum moving speed of the moving object; determining a moving distance according to the time interval and the maximum moving speed; for each sample, acquiring the abscissa position of the sample at the current moment according to the abscissa position of the sample at the previous moment, the moving distance, the sine value of the moving angle of the sample and the state transition model; acquiring the ordinate position of the sample at the current moment according to the ordinate position of the sample at the previous moment, the moving distance, the cosine value of the moving angle of the sample and the state transition model;

wherein the movement angle follows a uniform distribution of (0,2 π).

In one possible implementation, the apparatus further includes:

the sample resampling module is used for acquiring the effective sample number based on the weight of the sample at the current moment after acquiring the weight of the sample at the current moment for each sample; when the number of the obtained effective samples is smaller than the number threshold, resampling the samples;

wherein the probability of occurrence of the resampled sample at a location is proportional to the weight of the sample at the location prior to the resampling, and the weights are equal between different samples after the resampling.

In a possible implementation manner, the movement track generation module is further configured to filter the occurrence position data of the moving object at a preset number of moments; and connecting the appearance positions of the moving object at the rest of a plurality of moments according to the time sequence to obtain the moving track of the moving object.

In a possible implementation manner, the sample initialization module is further configured to perform multiple random sampling on the target geographic area based on two-dimensional uniform distribution; wherein the plurality of samples are equally weighted initially and are evenly distributed within the target geographic area.

In a possible implementation manner, the sample weight updating module is further configured to perform normalization processing on the weight of the generated sample at the current time.

In a possible implementation manner, the position estimation module is further configured to perform weighted average on the position information of the generated samples at the current time based on the weights of the generated samples at the current time, so as to obtain the appearance position of the mobile object in the target geographic area at the current time.

In another aspect, a storage medium is provided, where at least one instruction is stored, and the at least one instruction is loaded and executed by a processor to implement the movement track obtaining method described above.

In another aspect, an electronic device is provided, where the device includes a processor and a memory, where the memory stores at least one instruction, and the at least one instruction is loaded and executed by the processor to implement the movement trajectory acquisition method described above.

The technical scheme provided by the embodiment of the application has the following beneficial effects:

when the moving track of any moving object is obtained, sample initialization can be firstly carried out, namely, a target geographic area is sampled, and a plurality of samples representing possible positions of the moving object in the target geographic area are obtained; after sampling is completed, position estimation is periodically performed, that is, at each time of performing position estimation, the embodiment of the present application obtains the position information of a generated sample at the current time based on the position information of the generated sample at the previous time; in addition, updating the weight of the sample, namely acquiring the signal intensity matched with the identity of the terminal carried by the mobile object at the current moment, and further acquiring the weight of the generated sample at the current moment based on the acquired signal intensity and the path loss model; then, based on the position information and the weight of the generated sample at the current moment, the position of the moving object at the current moment can be determined; and repeatedly executing the position state updating, the weight updating and the position estimating process of the samples to obtain the appearance positions of the moving object at all times, and further connecting the appearance positions of the moving object at all times according to the time sequence to obtain the moving track of the moving object.

Based on the above description, when calculating the moving track of any moving object, the embodiment of the present application only needs to acquire the signal strength matched with the identity of the terminal carried by the moving object, and the position estimation of the moving object does not need to perform similarity comparison based on the signal strength, so that fingerprint information acquisition does not need to be performed at each reference position point in the target geographic area in advance.

Drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.

Fig. 1 is a schematic diagram of an implementation environment related to a movement trajectory acquisition method provided in an embodiment of the present application;

FIG. 2 is a diagram of a product effect of a moving track provided by an embodiment of the present application;

fig. 3 is a flowchart illustrating an overall execution of a method for obtaining a movement track according to an embodiment of the present disclosure;

fig. 4 is a flowchart of a method for acquiring a movement track according to an embodiment of the present application;

fig. 5 is a schematic view of a particle distribution after initialization of particles according to an embodiment of the present application is completed;

fig. 6 is a schematic view of a particle distribution after a state transition of particles according to an embodiment of the present application;

fig. 7 is a diagram illustrating an effect of updating a particle weight according to an embodiment of the present application;

fig. 8 is a schematic diagram of a particle distribution after particle resampling according to an embodiment of the present application;

fig. 9 is a schematic diagram of a location distribution after location estimation according to an embodiment of the present application;

FIG. 10 is a diagram of product effects of a movement track provided by an embodiment of the present application;

FIG. 11 is a diagram illustrating product effects of a moving track according to an embodiment of the present disclosure;

FIG. 12 is a diagram illustrating product effects of a moving track according to an embodiment of the present disclosure;

FIG. 13 is a diagram illustrating product effects of a moving track according to an embodiment of the present disclosure;

fig. 14 is a flowchart of a method for acquiring a movement track according to an embodiment of the present application;

fig. 15 is a schematic structural diagram of a movement trajectory acquisition device according to an embodiment of the present application;

fig. 16 is a schematic structural diagram of a movement trajectory acquisition device according to an embodiment of the present application.

Detailed Description

To make the objects, technical solutions and advantages of the present application more clear, embodiments of the present application will be described in further detail below with reference to the accompanying drawings.

Before explaining the embodiments of the present application in detail, abbreviations and key terms referred to in the embodiments of the present application are defined.

AP (Access Point): in general terms, an AP is a device that provides a connection between a wireless network and a wired network, i.e., an AP is an access point for a mobile terminal to enter the wired network.

In the embodiment of the present application, the AP mainly refers to a device capable of detecting a Media Access Control (MAC) address and signal strength of a mobile terminal that is close to a Wireless internet Access, such as a Wireless router and a Wireless Fidelity (WiFi) probe.

MAC address: the device identifier is a hardware address, and mobile terminals such as mobile phones can be identified through the MAC address. Wherein, a device usually corresponds to a MAC address; some vendors may employ a randomly generated MAC address for privacy reasons.

RSSI: refers to the signal strength before the transceiver device.

In the embodiment of the present application, the RSSI is used to indicate the signal strength between the AP and the mobile terminal.

Generally, the closer the distance the stronger the signal, the farther the distance the signal. Therefore, the distance between the AP and the mobile terminal can be inferred by the RSSI.

A WiFi probe: a device automatically identifies nearby devices with WiFi (wireless fidelity) functionality turned on based on WiFi detection technology. The WiFi probe can obtain data such as MAC addresses and RSSI values of nearby devices with WiFi functions turned on.

The device includes, but is not limited to, a smart phone, a notebook computer, a tablet computer, and the like. As one example, uses of WiFi probes include, but are not limited to, deployment within a store for passenger flow monitoring, and the like.

Log-distance Path Loss Model (Log-distance Path Loss Model): according to the characteristic that the radio propagates in free space, the energy of the received signal can be deduced to be positively correlated with the logarithm of the distance, wherein the logarithm distance path loss model is as follows:

wherein, the RSSI represents the signal strength value when the distance between the transceiver is d; RSSI₀To be at a close distance d₀A reference signal strength value of (c); as an example, d₀Is typically 1 meter; a is the path loss exponent, indicating the rate at which the path loss increases with distance, the value of which depends on the surrounding environment and building type.

Bayes' Theorem (Bayes): to describe the relationship between two conditional probabilities.

In another expression, the bayesian theorem is a theorem for calculating the posterior probability by using the prior probability and the likelihood, that is, the posterior probability (likelihood) is a prior probability/a normalized constant, and the calculation formula is as follows:

wherein, H is an assumption that probability needs to be calculated, D is observation data, and p (H | D) is a probability that H is established under the condition that data D is observed, i.e., a posterior probability; p (h) is the probability before data D is not considered, i.e. the prior probability; p (dl | H) is the probability that data D is observed under the H hypothesis, i.e., the likelihood probability; p (D) is the probability that data D is observed under any condition, which is a normalization constant.

Markov chain (Markov chain): a random process that is "memoryless", i.e. the probability distribution of the next state can only be determined by the current state, and events that precede the current state in the time series are all independent of the next state.

Particle Filtering (Particle Filtering): the Monte Carlo method, which uses a set of particles to represent probabilities, can be used on any form of state space model. The core idea is to express the distribution of random state particles by extracting the particles from the posterior probability, and the method is a sequential importance sampling method.

Briefly, the particle filtering method is a process of obtaining a state minimum variance distribution by finding a group of random samples propagating in a state space, approximating a probability density function, and substituting an integral operation by a sample mean. These samples are visually referred to as "particles" and are therefore referred to as particle filters.

Stated another way, particle filtering is a monte carlo method that estimates the state of a dynamic system from a noisy or incomplete observation sequence by representing the posterior probability of a random event by a set of weighted random samples.

An implementation environment related to the movement trajectory acquisition method provided by the embodiment of the present application is described below.

Referring to fig. 1, the implementation environment includes: an access point 101, a terminal 102 carried by a mobile object, and a device 103 for executing the movement trajectory acquisition method provided by the embodiment of the present application.

The number of the access points 101 is usually multiple, and the access points are deployed in a target geographic area, where the target geographic area may be any geographic area with a statistical passenger flow demand, such as a place with a large passenger flow volume, such as a shopping mall, and this is not specifically limited in this embodiment of the present application. It should be noted that, in the embodiment of the present application, the access point 101 mainly refers to a device capable of detecting an identity and a signal strength of a mobile terminal that is close to a wireless internet capable device, such as a wireless router and a WiFi probe.

As an example, the above identity may refer to a MAC address of the terminal in the embodiment of the present application.

For the terminal 102, it is typically a mobile terminal, of the type including, but not limited to, a smart phone, a tablet computer, and the like. As the mobile object carrying terminal 102 moves within the target geographic area, it can be detected by access points deployed at different locations.

In a possible implementation manner, the device 103 may be a server, and the device 103 is configured to collect the identity of the terminal detected by the access point 101 and the signal strength at the location of the detected terminal, and accordingly analyze a moving track of the moving object in the target geographic area, where fig. 2 shows a product effect diagram of the moving track. In the embodiment of the present application, the mobile object may be a pedestrian within the target geographic area.

As an example, the device 103 may push the analyzed movement tracks of the respective moving objects to other terminals for foreground display, for example, push the movement tracks of the guests to a terminal of the merchant in a form of a webpage, so that the merchant may view the movement tracks of different guests in the merchant.

In detail, the embodiment of the present application provides a method for analyzing a moving trajectory of a moving object based on a MAC address and signal strength detected by the access point 101. By the method, the moving track of the moving object in the target geographic area can be known, for example, the moving track of each user in a mall can be known, so that the purpose of knowing the interest of the user to optimize the user experience is achieved.

As shown in fig. 3, the embodiment of the present application mainly obtains the moving trajectory of the moving object through the following processes:

a. after the access point detects the raw data, the detected raw data is subjected to data preprocessing.

In the embodiment of the present application, the raw data is a general term for the MAC address and the signal strength detected by the access point.

b. And carrying out parameter estimation on the path loss model. The path loss model refers to the aforementioned logarithmic distance path loss model, and this step is to estimate the reference signal strength and the path loss exponent in the path loss model.

It should be noted that, the order of executing the steps a and b and the steps c and d described below is arbitrary, and this is not specifically limited in this embodiment of the present application.

c. Particle initialization is performed. In the embodiment of the present application, the particles are used to characterize the possible occurrence positions of the mobile object in the target geographic area.

d. And performing state transition of the particles based on the state transition model.

e. And updating the particle weight based on Bayesian theorem, and resampling the particles.

f. And carrying out position estimation on the moving object by combining a particle filtering algorithm.

g. And connecting the estimated positions to obtain the moving track of the moving object in the target geographic area.

Based on the above description, as the moving object moves in the target geographic area, it may be detected by the access points installed and deployed at different locations. Since the MAC address of the terminal carried by the same moving object is usually not changed, in the embodiment of the present application, the same MAC address detected by each access point can be used as the identity of the moving object carrying the corresponding terminal, and then, in combination with the location deployment information of each access point, when the access points at different locations detect the same terminal, the detected signal strengths are different, and thus, the moving trajectory of the moving object can be calculated.

In summary, the method for acquiring the movement track provided by the embodiment of the present application does not need to acquire fingerprint information at each reference location point in the target geographic area in advance, so that offline workload is small, fast multiplexing and large-scale expansion are supported, universality is good, in addition, no additional operation needs to be performed after deployment change in the target geographic area, and an effect is good. In addition, the method is implemented only by deploying the access points in the target geographic area, so that the cost is low, and the WiFi signals can penetrate through the shelters such as walls to a certain extent, so that the requirements on the shelters are not strict, and the mobile track acquisition can be well supported. In addition, the method is convenient to realize tracking the same moving object, and only the same MAC address needs to be identified. In short, in the embodiment of the present application, under the condition that fingerprint information does not need to be acquired offline, the moving tracks of different moving objects are acquired based on the execution flowchart shown in fig. 3 through MAC addresses and signal strengths detected by access points deployed at different locations.

The following explains a movement trajectory acquisition method provided in the embodiment of the present application in detail.

It should be noted that the descriptions like first, second, third, fourth, etc. appearing below are only for the purpose of distinguishing different objects, and do not constitute any other limitation. In addition, based on the foregoing, the particles may also be referred to as samples herein, in other words, the particles and the samples have a uniform meaning and are used to characterize the possible positions of the mobile object in the target geographic area.

Fig. 4 is a flowchart of a method for acquiring a movement track according to an embodiment of the present application. The execution subject of the method may be the apparatus 103 shown in fig. 1. Referring to fig. 4, a method flow provided by the embodiment of the present application includes:

401. raw data detected by various access points deployed within a target geographic area is acquired.

The original data includes the identity of the terminal detected by each access point, the signal strength at the position where the terminal is detected, and a timestamp. Wherein a terminal as referred to herein generally refers to a mobile terminal. Generally, one terminal corresponds to one moving object.

In this embodiment of the application, the target geographic area is usually an indoor geographic area, and taking the indoor geographic area as a mall and the identity of the terminal as the MAC address of the terminal as an example, each AP deployed in the mall may detect a nearby mobile terminal having a WiFi function turned on, and obtain data such as the detected MAC address and signal strength value of the mobile terminal. In order to distinguish the data before and after performing the data preprocessing, the embodiment of the present application refers to the data before the preprocessing as the original data.

In one possible implementation, the original data is a set of (ts, end _ mac, rssi) data. Wherein ts is a timestamp used for indicating the detection time, endpoint _ MAC indicates the MAC address of the detected mobile terminal, and rsi is the signal strength at the position of the detected mobile terminal.

As shown in table 1 below, one AP may detect one or more mobile terminals at the same time, for example, AP1 detects both mobile terminal Endpoint1 and mobile terminal Endpoint 1; in addition, the same mobile terminal may be detected by multiple APs at the same time, for example, AP1, AP2, and AP3 all detect mobile terminal Endpoint1 at the same time.

TABLE 1

402. And performing data preprocessing on the detected raw data.

In one possible implementation, the data pre-processing includes at least one of:

firstly, redundant data in original data are filtered.

Wherein the redundant data includes at least abnormal data and data unrelated to the moving object. As an example, this step is used to filter out anomalous data in the raw data as well as other data that is not relevant to the passenger flow. For example, the original data acquired when no person is in a shopping mall in the morning belongs to other data irrelevant to passenger flow, because the data acquired in the period is likely to be data of other household appliances, not data of a mobile terminal carried by the person in the shopping mall.

And secondly, dividing the signal intensity data included in the original data according to the detected identity and the time stamp of the terminal to form a corresponding relation with the identity and the time stamp as indexes and the signal intensity data as values.

Because the detection frequencies of different APs are not necessarily synchronous, the original data acquired by different APs also needs to be aligned according to the timestamp and the MAC address of the mobile terminal.

In a possible implementation manner, as shown in table 2 below, in the embodiment of the present application, original data acquired by each AP is aggregated into a table with (ts, end _ mac) as an index and RSSI values of mobile terminals acquired by each AP as a column.

In addition, in the implementation of the present application, data belonging to the same mobile terminal are grouped into one group, that is, after the step is performed, each group of data in the following table 2 corresponds to one pedestrian and a moving track of the pedestrian.

TABLE 2

403. And carrying out parameter estimation on the path loss model.

As an example, the path loss model refers to the aforementioned logarithmic distance path loss model, and this step is to estimate the reference signal strength and the path loss exponent in the path loss model. In another expression, this step is used to obtain the values of the reference signal strength and the path loss index in the path loss model.

In the embodiment of the present application, the values of the reference signal strength and the path loss exponent are obtained, including but not limited to the following two ways:

4031. and when the distance between the access points deployed in the target geographic area is smaller than a distance threshold, acquiring signal intensity data collected when the access points perform mutual detection, and fitting values of reference signal intensity and a path loss index according to the distance between the access points and the collected signal intensity data.

The method aims at the condition that the AP deployment in the target geographic area is dense, namely when the AP deployment distance in the target geographic area is dense, the AP can be mutually detected, and then RSSI data can be collected. And then, the distances among the APs and the corresponding RSSI values are calculated through a plane, and the values of the reference signal strength and the path loss index can be fitted by combining a least square method.

The distance threshold may be 5 meters or 10 meters, and the like, which is not specifically limited in the embodiment of the present application.

4032. When the distance between each access point is greater than a distance threshold, acquiring signal intensity data acquired by detection equipment when detecting each access point at different positions; values of reference signal strength and path loss exponent are obtained based on signal strength data collected at different locations.

This kind of mode is to the condition that AP deployment is comparatively sparse in the target geographic area, promptly when can not surveying each other between each AP in the target geographic area, can hold the detection equipment by the staff in advance and survey each access point in different positions, gathers RSSI data, and then based on the RSSI data that gathers in different positions, obtains the value of reference signal intensity and path loss index.

As an example, for this case, the maximum RSSI value and the minimum RSSI value that can be detected by the APs may be assumed, such as setting the maximum RSSI value detected by an AP to be the RSSI value at 1 meter or at the nearest grid point, and the minimum RSSI value detected to be the RSSI value at the farthest position (for example, at 20 meters) of the detection range, and then performing parameter estimation based on the maximum RSSI value and the minimum RSSI value detected.

In the embodiment of the present application, the estimated RSSI is used₀Taking the value of b and the path loss exponent as an example, then according to the path loss model, the RSSI value when the terminal with the distance d from the AP is detected can be obtained: RSSI ═ axlogd + b; similarly, the distance between the AP and the terminal can be reversely deduced according to the path loss model based on the RSSI value detected by the AP:

it should be noted that, the sequence of the steps 401 to 403 and the

steps

404 and 405 described below is arbitrary, and this is not specifically limited in this embodiment of the present application.

404. Particle initialization is performed on the target geographic area.

In the embodiment of the present application, the particle initialization process is to sample a target geographic area, and to refer to the sampled sample as a particle, to obtain a plurality of particles for characterizing a possible occurrence position of a moving object in the target geographic area.

In detail, the embodiment of the application performs particle initialization on a two-dimensional plane of a target geographic area.

In one possible implementation, the target geographic area is sampled multiple times, including but not limited to: and carrying out multiple random sampling on the target geographic area based on two-dimensional uniform distribution. That is, in the initial case, the present embodiment considers that the probability of the mobile object appearing at any position in the target geographic area is equal.

As an example, referring to fig. 5, the embodiment of the present application performs N random samplings with a two-dimensional uniform distribution. Wherein, the value of N is a positive integer, and the generated particles are uniformly distributed in the plane of the target geographic area.

In addition, the initial weights of the generated particles are equal, that is, the weight of the particle generated in the initial case is equal, and it is assumed that the weight of the ith particle is w_iDue to the fact that

Thus, it is possible to provide

In fig. 5, the dark color point is the position of the AP, and the light color point is the sampling particle.

405. A state transition is performed on the generated particles.

In each iteration cycle in time, the embodiment of the present application performs a round of particle state transition process. Each iteration period in time refers to a time period during which a position estimate is made.

In one possible implementation, assuming that the position state of the moving object is based on a markov chain, i.e. the position of the moving object at the next time instant is related only to the position of its current time instant and not to the positions that have occurred in the past, the state transition equation for the moving object can be expressed as:

x_k,y_k＝f_k(x_k-1,y_k-1)

wherein (x)_k,y_k) Refers to the position coordinates of the particle k at the moment in the plane of the target geographical area, f_kDenotes the transition transfer function, (x)_k-1,y_k-1) Refers to the position coordinates of particle k-1 in the plane of the target geographic area at the time. As can be seen from the above description, for the current time, the position information of the generated particle at the current time is acquired based on the position information of the generated particle at the previous time.

Taking the target geographic area as the mall as an example, then (x)_k,y_k) I.e. the position coordinate of the particle k in the mall plane at the moment. As described above, such an equation that functionally expresses the relationship between the front and rear stages is referred to as a state transition equation. In addition, the above state transition equation may also be referred to as a state transition model.

For the state transition process, obtaining a state transition model by using another expression; and acquiring the position information of the generated particle at the current moment based on the position information of the generated particle at the previous moment and the state transition model. As an example, the position information of the generated particle at the current time is obtained based on the position information of the generated particle at the previous time and the state transition model, which includes but is not limited to:

acquiring the time interval between the previous moment and the current moment, and acquiring the maximum moving speed of the moving object; determining a moving distance according to the time interval and the maximum moving speed; for each particle, acquiring the abscissa position of the particle at the current time according to the abscissa position of the particle at the previous time, the determined movement distance, the sine value of the movement angle of the particle and the state transfer function; and acquiring the vertical coordinate position of the particle at the current moment according to the vertical coordinate position of the particle at the previous moment, the moving distance, the cosine value of the moving angle of the particle and the state transfer function.

As an example, assuming that the maximum moving speed of the moving object is v (m/s), the possible positions of the moving object after the time interval Δ t between the previous time and the current time are (x)_k-1,y_k-1) A circular area with v delta t as a radius and used as a circle center. In the actual calculation, an area blocked by an obstacle such as a wall should be removed.

In the embodiment of the present application, assuming that the movement of the moving object is random, the movement moves to any point in the circular area with the same probability, and thus the movement moves to any point in the circular area

x_k＝x_k-1+d*sinθ

y_k＝y_k-1+d*cosθ

Wherein, theta to Unif (0,2 pi), d to Unif (0, v delta t).

It should be noted that when the particle filter is applied in the fields of robot or automation control, the state transition equation is usually used to describe the state change of the robot or machine after being subjected to the control command or calculated according to the dynamics. In the embodiment of the present application, the state transition equation more greatly acts to limit the possible movement range of the mobile object due to the unpredictability of the movement of the mobile object.

In the embodiment of the present application, the position of the generated particle can be updated by the above formula in each iteration cycle in time. See fig. 6, which shows the position of the particles after the state transition has occurred.

406. The generated particles are weight updated and resampled.

In a possible implementation manner, the embodiment of the present application updates the particle weight based on bayesian theorem. In detail, a round of particle weight update and resampling process is performed in each iteration cycle in time.

Aiming at the weight updating process at the current moment, namely acquiring the signal intensity matched with the identity of the terminal carried by the mobile object at the current moment, and updating the weight of the generated particles at the current moment based on the acquired signal intensity and the path loss model; the obtaining of the signal strength matching the identity of the terminal carried by the mobile object at the current moment includes, but is not limited to: and acquiring the signal intensity matched with the identity of the terminal carried by the mobile object at the current moment from the preprocessed original data.

In one possible implementation, updating the weight of the generated particle at the current time based on the obtained signal strength and the path loss model includes the following steps:

4061. for each particle, the weight of the particle at the last time instant is obtained.

4062. Acquiring the signal intensity detected by each access point at the position of the particle from the acquired signal intensity; based on the path loss model, likelihood probabilities of signal strengths detected at locations where the particle is located are obtained.

Assuming that the current time is k times, the observed value of the signal strength detected by an AP at the k times is RSSI_kFor example, the posterior probability can be obtained from the prior probability by using bayesian theorem as shown in the following formula:

wherein z is_k＝(x_k，y_k) Refers to the position coordinates of the particle at time k in the plane of the target geographic area, RSSI_kThe signal strength value detected by an AP at the location specified by the location coordinates at time k.

In addition, in the above formula, p (z)_k|RSSI_k) For a posteriori probability, refer to the measured RSSI_kUnder the condition of Z_kA probability of being established; p (z)_k) For a priori probability, refer toRSSI filtering_kThe probability of the previous; p (RSSI)_k|z_k) For likelihood probability, refer to at Z_kDetecting RSSI under established conditions_kThe probability of (d);

to normalize constant, refer to the RSSI detected under any condition_kThe probability of (c).

It should be noted that, since each access point may possibly detect at the position of the particle, likelihood probabilities of the signal strengths detected at the position of the particle may be respectively calculated based on the signal strengths detected at the position of the particle by each AP, and finally, a plurality of obtained likelihood probability results are combined to obtain a final likelihood probability.

4063. Acquiring the weight of the particle at the current moment based on the weight of the particle at the previous moment and the likelihood probability; wherein the weight of the particle at the current time instant is proportional to the likelihood probability.

As an example, the position coordinates of the ith randomly sampled sample (i.e., particle) at time k is taken as

Corresponding weight is

Can obtain

Is to be used to represent the frequency of the,

in view of the above formula, the embodiment of the present application needs to calculate p (RSSI)_k|Z_k ⁱ) I.e. the likelihood probability that the RSSI value measured at a certain position needs to be calculated.

As an example, assuming that the distribution of the likelihood probability is a normal distribution, the expected value of the normal distribution, that is, the likelihood probability obeys the normal distribution, can be obtained by the above-mentioned path loss model, and the expected value of the normal distribution is obtained based on the above-mentioned path loss model RSSI ═ a × logd + b.

That is, the following formula for calculating the likelihood probability can be obtained according to the specified standard deviation σ:

p(RSSI＝rssi)～N(-A*logd+b,σ)

the standard deviation σ may be obtained through an experiment, and d in the formula may be calculated according to the deployment position of the AP and the position of the particle, which is not specifically limited in this embodiment of the present application.

In addition, due to

Therefore, after the particle weight is updated in each iteration period in time, normalization needs to be performed again, wherein the normalization result is

Fig. 7 shows an effect diagram after updating the weight of the particle, and in fig. 7, the size of the particle is proportional to the size of the weight.

In the embodiment of the present application, a degradation problem (degradation problem) may be caused in the particle weight updating process, that is, the weights of many particles are reduced to be negligible after several iterations, so that the calculation performance is reduced. In one possible implementation, the resampling process includes, but is not limited to:

for each particle, after the weight of the particle at the current moment is obtained, the number of effective particles is obtained based on the weight of the particle at the current moment; when the number of the obtained effective particles is smaller than the number threshold, performing primary particle resampling; wherein the probability of occurrence of the resampled particle at a position is proportional to the weight of the particle at the position before the resampling, and the weights are equal between different resampled particles.

As an example, for the particle resampling process, the current number of valid particles needs to be calculated based on the following formula.

Wherein w denotes the weight of the particle, and the resampling principle may be: when the value of Neff is smaller than the weight threshold, for example, Neff < N/2, resampling is performed according to the weight of the original particle, that is, the probability of the resampled particle at a certain position is proportional to the weight of the original particle at the position. After resampling, the generated particle weight is required to be 1/N again. Where N denotes the number of particles generated during particle initialization.

Wherein fig. 8 shows the particle distribution after resampling.

407. And determining the appearance position of the mobile object in the target geographic area at the current moment based on the generated position information and the weight of the particles at the current moment.

This step is to perform position estimation on the moving object, and it should be noted that, in each iteration cycle in time, a round of position estimation process is performed.

In one possible implementation, the position of the mobile object in the target geographic area at the current time is determined based on the position information and the weight of the generated particle at the current time, which includes but is not limited to: and carrying out weighted average on the position information of the generated particles at the current moment based on the weight of the generated particles at the current moment to obtain the appearance position of the moving object in the target geographic area at the current moment.

Wherein the calculation formula for position estimation is

Wherein the content of the first and second substances,

refers to the estimated position of occurrence, w, of the moving object at the current time_kWeight, z, referring to the particle_kRefers to the position coordinates of the particle. In the case of performing the position estimation, the position coordinates of all the particles are weighted and averaged according to the weight.

Fig. 9 shows an effect diagram of performing position estimation, and in fig. 9, a black dot is an estimated position of the moving object.

408. And connecting the appearance positions of the mobile objects in the target geographic area at various moments according to the time sequence to obtain the moving track of the mobile objects.

In this step, the movement locus of the moving object is formed. After the position estimation under each time period is completed, connecting a plurality of results after the position estimation into a line according to the time sequence to obtain the moving track of the moving object.

In one possible implementation, since the first few time periods are greatly affected by the initialization conditions, the finally generated movement trajectory may discard the position estimation results of the previous several time periods. That is, connecting the appearance positions of the mobile object in the target geographic area at each moment in time sequence to obtain the movement track of the mobile object, including: filtering the appearance position data of the mobile object at the previous preset number of moments; and connecting the appearance positions of the moving object at the rest of the plurality of moments according to the time sequence to obtain the moving track of the moving object. Among them, fig. 10 to 13 show the effect diagrams of the movement locus.

The method provided by the embodiment of the application at least has the following beneficial effects:

when the moving track of any moving object is calculated, only the signal intensity matched with the identity of the terminal carried by the moving object needs to be acquired, the position of the moving object does not need to be estimated and compared based on the similarity of the signal intensity, so that fingerprint information collection does not need to be carried out on each reference position point in a target geographic area in advance.

In another expression, the moving tracks of different moving objects can be acquired by the MAC addresses and the signal strengths detected by the access points deployed at different positions in the target geographic area. The method does not need to acquire the fingerprint information at each reference position point in the target geographic area in advance, so that the offline workload is small, the method is easy to realize, quick multiplexing and large-scale expansion are supported, the universality is good, no additional operation is needed after the deployment in the target geographic area is changed, and the effect is good.

In addition, the method is implemented only by deploying the access points in the target geographic area, so that the cost is low, and the WiFi signals can penetrate through the shelters such as walls to a certain extent, so that the requirements on the shelters are not strict, and the mobile track acquisition can be well supported. In addition, the method is convenient to realize tracking the same moving object, and only the same MAC address needs to be identified.

In addition, the embodiment of the application also provides a state transition model suitable for moving of the mobile object, which is used for limiting the range to which the mobile object may move within a period of time, and estimating the moving track of the mobile object within a period of continuous time by combining a particle filtering method, so that habits of different people can be known through the moving track, for example, taking a target geographic area as a shopping mall as an example, and by knowing the moving track of different people in the shopping mall, shopping habits of different users can be quickly known, and a good foundation is laid for providing personalized services for different users.

In another embodiment, an embodiment of the present application provides a method for acquiring a movement trajectory, and referring to fig. 14, a flow of the method provided by the embodiment of the present application includes:

1401. and sampling the target geographic area to obtain a plurality of samples representing the possible positions of the mobile object in the target geographic area.

This step is the same as step 404 described above.

1402. And at each moment when the position of the moving object is estimated after the sampling is finished, acquiring the position information of the generated sample at the current moment based on the position information of the generated sample at the previous moment.

This step is the same as step 405 above.

1403. And acquiring the signal intensity matched with the identity of the terminal carried by the mobile object at the current moment, and acquiring the weight of the generated sample at the current moment based on the acquired signal intensity and the path loss model.

This step is the same as the process of updating the weights of the generated samples in step 406.

1404. And determining the appearance position of the mobile object in the target geographic area at the current moment based on the position information and the weight of the generated sample at the current moment.

This step is the same as step 407 described above.

1405. And connecting the appearance positions of the mobile objects in the target geographic area at various moments according to the time sequence to obtain the moving track of the mobile objects.

This step is the same as step 408 described above.

According to the method provided by the embodiment of the application, when the moving track of any moving object is calculated, only the signal intensity matched with the identity of the terminal carried by the moving object needs to be acquired, the similarity comparison based on the signal intensity is not needed for the position estimation of the moving object, and therefore fingerprint information collection does not need to be carried out on each reference position point in a target geographic area in advance. In another expression, in the embodiment of the present application, the moving tracks of different moving objects can be obtained by using the identity and the signal strength detected by the access points deployed at different positions in the target geographic area. The method does not need to acquire the fingerprint information at each reference position point in the target geographic area in advance, so that the offline workload is small, the method is easy to realize, quick multiplexing and large-scale expansion are supported, the universality is good, no additional operation is needed after the deployment in the target geographic area is changed, and the effect is good.

In addition, the method is implemented only by deploying the access points in the target geographic area, so that the cost is low, and the WiFi signals can penetrate through the shelters such as walls to a certain extent, so that the requirements on the shelters are not strict, and the mobile track acquisition can be well supported. In addition, the method is convenient to realize tracking the same moving object, and only the same identity identification needs to be recognized.

In addition, after the movement track of the mobile object in a period of continuous time is estimated, the habits of different people can be known through the estimated movement track, for example, taking a target geographic area as a shopping mall as an example, and the shopping habits of different users can be quickly known through knowing the movement track of different people in the shopping mall, so that a good foundation is laid for providing personalized services for different users.

In another embodiment, taking the target geographic area as a shopping mall and the moving object as a pedestrian in the shopping mall as an example, an overall execution flow of the movement track obtaining method provided by the embodiment of the present application is described with reference to fig. 3.

a. The method comprises the steps of acquiring raw data detected by access points deployed at different positions in a shopping mall, and preprocessing the detected raw data.

b. And carrying out parameter estimation on the path loss model.

The path loss model refers to the aforementioned logarithmic distance path loss model, and this step is to estimate the reference signal strength and the path loss exponent in the path loss model.

c. And carrying out particle initialization on the plane of the market. In this case, the particles are used in the exemplary embodiment of the present application to characterize the possible positions of pedestrians in the plane of the store.

d. Based on the state transition model, a particle state transition is performed on the generated particle.

e. And updating the weight of the generated particles based on Bayesian theorem, and resampling the particles.

f. The position of the pedestrian in the plane of the mall is estimated.

g. And connecting the estimated positions to obtain the moving track of the pedestrian in the market.

After the moving track of the pedestrian in a continuous time is estimated, the habits of different people can be known through the estimated moving track, in the embodiment, the shopping habits of different users can be quickly known through knowing the moving tracks of the different people in a shopping mall, and a good foundation is laid for providing personalized services for the different users. The personalized service includes, but is not limited to, pushing fashion information or discount message related to shopping habits of the user to the user, and the embodiment of the present application is not particularly limited thereto.

In another embodiment, in addition to the previously enumerated embodiments, the present examples include the following alternative embodiments.

In the foregoing embodiment, the state transition model of the moving object defines the moving range of the moving object within a time interval as a circular region, and the probability of the moving object reaching each point within the circular region is equal. In addition to this, it is also possible to model the position state and motion of a moving object with a more accurate model, e.g. adding the velocity and orientation of the moving object to the estimation of the position state, since in an actual scene the probability of the moving object moving in the same direction will generally be higher than the probability of turning.

In the foregoing embodiment, the initialization particles are randomly sampled based on two-dimensional uniform distribution, and in an actual application process, if there is a clear entry in the target geographic area, the particles may be initialized according to the actual entry.

In the foregoing embodiment, the RSSI likelihood probability obeys normal distribution, and in addition, the RSSI likelihood probability also obeys log normal distribution, or actual measurement data is adopted to fit corresponding distribution, which is not specifically limited in this embodiment of the application.

Fig. 15 is a schematic structural diagram of a movement trajectory acquisition device according to an embodiment of the present application. Referring to fig. 15, the apparatus includes:

the sample initialization module 1501 is configured to sample a target geographic area to obtain multiple samples representing possible positions of a mobile object in the target geographic area;

a sample position updating module 1502, configured to obtain, at each time of performing position estimation on the moving object after completing sampling, position information of the generated sample at the current time based on the position information of the generated sample at the previous time;

a sample weight updating module 1503, configured to obtain, at each time of performing position estimation on the moving object after completion of sampling, a signal strength at the current time that matches an identity of a terminal carried by the moving object, and obtain a weight of the generated sample at the current time based on the obtained signal strength and the path loss model;

a position estimation module 1504, configured to determine, at each time of performing position estimation on the mobile object after completing sampling, an appearance position of the mobile object in the target geographic area at the current time based on the position information and the weight of the generated sample at the current time;

the moving track generating module 1505 is configured to connect, according to a time sequence, appearance positions of the mobile objects in the target geographic area at various times to obtain a moving track of the mobile objects.

According to the device provided by the embodiment of the application, when the moving track of any moving object is obtained, sample initialization can be firstly carried out, namely, a target geographical area is sampled, and a plurality of samples representing the possible positions of the moving object in the target geographical area are obtained; after sampling is completed, position estimation is periodically performed, that is, at each time of performing position estimation, the embodiment of the present application obtains the position information of a generated sample at the current time based on the position information of the generated sample at the previous time; in addition, updating the weight of the sample, namely acquiring the signal intensity matched with the identity of the terminal carried by the mobile object at the current moment, and further acquiring the weight of the generated sample at the current moment based on the acquired signal intensity and the path loss model; then, based on the position information and the weight of the generated sample at the current moment, the position of the moving object at the current moment can be determined; and repeatedly executing the position state updating, the weight updating and the position estimating process of the samples to obtain the appearance positions of the moving object at all times, and then connecting the appearance positions of the moving object at all times according to the time sequence to obtain the moving track of the moving object.

In one possible implementation, the apparatus further includes:

In one possible implementation, the apparatus further includes;

wherein the movement angle follows a uniform distribution of (0,2 π).

In one possible implementation, the apparatus further includes:

All the above optional technical solutions may be combined arbitrarily to form the optional embodiments of the present disclosure, and are not described herein again.

It should be noted that: in the above embodiment, when the movement trace obtaining apparatus obtains the movement trace, only the division of the functional modules is taken as an example, and in practical applications, the function distribution may be completed by different functional modules according to needs, that is, the internal structure of the apparatus is divided into different functional modules, so as to complete all or part of the functions described above. In addition, the embodiment of the movement track acquisition device and the embodiment of the movement track acquisition method provided by the above embodiments belong to the same concept, and specific implementation processes thereof are detailed in the method embodiments and are not described herein again.

Fig. 16 is a schematic structural diagram of an electronic device 1600 according to an embodiment of the present application, where the electronic device 1600 may generate a relatively large difference due to a difference in configuration or performance, and may include one or more processors (CPUs) 1601 and one or more memories 1602, where at least one instruction is stored in the memory 1602, and the at least one instruction is loaded and executed by the processors 1601 to implement the movement trajectory obtaining method provided by the foregoing method embodiments. Of course, the electronic device may further have components such as a wired or wireless network interface, a keyboard, and an input/output interface, so as to perform input/output, and the electronic device may further include other components for implementing the functions of the device, which is not described herein again.

In an exemplary embodiment, a computer-readable storage medium, such as a memory including instructions executable by a processor in a terminal, to perform the movement trace obtaining method in the above-described embodiment is also provided. For example, the computer readable storage medium may be a ROM, a Random Access Memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, and the like.

It will be understood by those skilled in the art that all or part of the steps for implementing the above embodiments may be implemented by hardware, or may be implemented by a program instructing relevant hardware, where the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a magnetic disk or an optical disk, etc.

The above description is only exemplary of the present application and should not be taken as limiting the present application, as any modification, equivalent replacement, or improvement made within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. A method for acquiring a movement track is characterized by comprising the following steps:

acquiring the signal intensity matched with the identity of the terminal carried by the mobile object at the current moment, and acquiring the weight of the generated sample at the current moment based on the acquired signal intensity and a path loss model;

determining the appearance position of the mobile object in the target geographic area at the current moment based on the position information and the weight of the generated sample at the current moment;

connecting the appearance positions of the mobile object in the target geographic area at various moments according to a time sequence to obtain a moving track of the mobile object;

the method further comprises the following steps: obtaining the values of the reference signal strength and the path loss index in the path loss model;

obtaining values of the reference signal strength and the path loss exponent includes:

when the distance between each access point deployed in the target geographic area is smaller than a distance threshold, acquiring signal intensity data collected when each access point mutually detects, and fitting values of the reference signal intensity and the path loss index according to the distance between each access point and the collected signal intensity data; or the like, or, alternatively,

when the distance between the access points is larger than the distance threshold, acquiring signal intensity data acquired by detection equipment when detecting the access points at different positions; obtaining values of the reference signal strength and the path loss exponent based on the signal strength data collected at different locations.

2. The method according to claim 1, wherein the obtaining the signal strength matching the identity of the terminal carried by the mobile object at the current time comprises:

acquiring raw data detected by each access point deployed in the target geographic area, wherein the raw data comprises an identity of a detected terminal and signal strength of the position of the detected terminal, and one terminal corresponds to one moving object;

and acquiring the signal intensity matched with the identity of the terminal carried by the mobile object at the current moment from the detected original data.

3. The method of claim 2, further comprising:

carrying out data preprocessing on the detected original data;

the obtaining of the signal strength matching the identity of the terminal carried by the mobile object at the current moment includes: acquiring the signal intensity matched with the identity of the terminal carried by the mobile object at the current moment from the preprocessed original data;

4. The method according to any one of claims 1 to 3, wherein the obtaining weights of the generated samples at the current time based on the obtained signal strength and path loss models comprises:

for each sample, acquiring the weight of the sample at the last moment;

acquiring the signal strength detected by each access point at the position of the sample from the acquired signal strength;

obtaining likelihood probabilities of signal strengths detected at locations where the samples are based on the path loss model;

acquiring the weight of the sample at the current moment based on the weight of the sample at the previous moment and the likelihood probability;

5. The method according to any one of claims 1 to 3, wherein the obtaining the position information of the generated sample at the current time based on the position information of the generated sample at the previous time comprises:

acquiring a state transition model;

and acquiring the position information of the generated sample at the current moment based on the position information of the generated sample at the previous moment and the state transition model.

6. The method according to claim 5, wherein the obtaining the position information of the generated sample at the current time based on the position information of the generated sample at the previous time and the state transition model comprises:

acquiring the time interval between the previous moment and the current moment, and acquiring the maximum moving speed of the moving object;

determining a moving distance according to the time interval and the maximum moving speed;

for each sample, acquiring the abscissa position of the sample at the current moment according to the abscissa position of the sample at the previous moment, the moving distance, the sine value of the moving angle of the sample and the state transition model;

acquiring the ordinate position of the sample at the current moment according to the ordinate position of the sample at the previous moment, the moving distance, the cosine value of the moving angle of the sample and the state transition model;

wherein the movement angle follows a uniform distribution of (0,2 π).

7. The method according to any one of claims 1 to 3, further comprising:

for each sample, after the weight of the sample at the current moment is obtained, obtaining the number of effective samples based on the weight of the sample at the current moment;

when the number of the obtained effective samples is smaller than the number threshold, resampling the samples;

8. The method according to any one of claims 1 to 3, wherein the connecting the appearance positions of the mobile object in the target geographic area at the respective moments in time in the time sequence to obtain the moving track of the mobile object comprises:

filtering the appearance position data of the mobile object at a preset number of moments;

and connecting the appearance positions of the moving object at the rest of a plurality of moments according to the time sequence to obtain the moving track of the moving object.

9. The method of any one of claims 1 to 3, wherein sampling the target geographic area comprises:

performing multiple random sampling on the target geographic area based on two-dimensional uniform distribution;

wherein the plurality of samples are equally weighted initially and are evenly distributed within the target geographic area.

10. The method according to any one of claims 1 to 3, further comprising:

and carrying out normalization processing on the weight of the generated sample at the current moment.

11. The method according to any one of claims 1 to 3, wherein the determining the appearance position of the mobile object in the target geographic area at the current time based on the position information and the weight of the generated sample at the current time comprises:

and carrying out weighted average on the position information of the generated samples at the current moment based on the weight of the generated samples at the current moment to obtain the appearance position of the mobile object in the target geographic area at the current moment.

12. A movement trajectory acquisition apparatus, characterized in that the apparatus comprises:

the mobile track generation module is used for connecting the appearance positions of the mobile objects in the target geographic area at all times according to the time sequence to obtain the mobile track of the mobile objects;

13. A storage medium having stored therein at least one instruction, which is loaded and executed by a processor to implement the movement trace obtaining method according to any one of claims 1 to 11.

14. An electronic device, comprising a processor and a memory, wherein the memory stores at least one instruction, and the at least one instruction is loaded and executed by the processor to implement the movement trace obtaining method according to any one of claims 1 to 11.