CN104469936A - Hybrid location method and system for wireless signal attenuation model based on intelligent space - Google Patents

Abstract

The invention relates to a hybrid location method and system for a wireless signal attenuation model based on intelligent space and belongs to the technical field of intelligent wheelchair location in the intelligent space. Aiming at the problem that due to random shaking and drifting of RSSI values collected in the wireless signal attenuation model, the locating precision is low, a Gauss filtering and averaging method is used for improving the accuracy of the collected RSSI values, and the problem that due to random shaking of RSSI, the range error is too large is solved to a certain extent; as a locating algorithm only based on the RSSI and arranged on hardware is single and has certain limitation, to improve the algorithm, an LQI auxiliary locating algorithm supported by multivariate is put forward, by means of the improved RSSI, the LQI hybrid locating algorithm, and a weighting dynamic optimization method, the range error measured finally is greatly reduced, then the indoor locating precision is improved, and the indoor locating system of the intelligent space can be applied to more application fields.

Description

Based on hybrid locating method and the system of the radio signal attenuation model of intelligent space

Technical field

The invention belongs to the intelligent wheel chair field of locating technology of intelligent space, relate to a kind of hybrid locating method and system of the radio signal attenuation model based on intelligent space.

Background technology

Along with the progress of society, science and technology is also in continuous development, and the life stress of society people is also in continuous reinforcement, and some the elderlys and physical disabilities often have selected and live by oneself to alleviate the burden brought to household.To the research of intelligent wheel chair and intelligent living space, safety and human nature service easily can not only be brought for the elderly and physical disabilities, guardian can also be allowed to understand their weather in time.Therefore very real meaning is had to the research of intelligent wheel chair intelligent space navigation system.

Intelligent space is the space system under multi information Cross-environment, and its research has been dissolved in the middle of daily life gradually.It has often been said that the Internet changes our life, the intelligent residence space system so comprising supervisory control system has then overturned our definition for family, makes people outside also can understand situation in family whenever and wherever possible, has deepened and the contacting of family.In addition, intelligent space and Internet of Things Robotics combine and can also be applied to each large-scale Public place.The each large well-known research institute in the whole world and giant company are all launching research to it, and propose multiple solution, and these will bring a new technological revolution.

In informationalized society, radio network technique is at development, and the requirement of wireless location technology is more and more higher, and application is also more and more wider.Location technology under the spacious environment in current outdoor is ripe, but the research of indoor positioning technologies is at the early-stage.It is not good that a lot of technology is used in indoor positioning effect, cannot reach indoor high-precision location requirement, so the exploitation and design of indoor locating system also result in the concern of people.Supermarket, airport, the large-scale public places such as home for destitute all need to rely on indoor positioning information realize management efficiently and improve better service.In the medical wireless communication technology, be also current important research direction to the intelligent monitor of patient and the follow-up service of position.

Radio sensing network is concentrated-distributed signals collecting, and the network information system of Signal transmissions and signal transacting and one, it is described as " fourth industrial revolution " of IT technology.This network is the core network of the intelligent space fixer network of this subject study, so have employed just in WSN technology based on the navigation system of the intelligent wheel chair of intelligent space.When intelligent wheel chair at indoor moving time obtain and the information of environmental information and physiological health, its spatial positional information can also be obtained simultaneously.

Intelligent wheel chair, as/disabled aiding robot of helping the elderly, needs its modularization, intellectuality, personalization and multi-functional etc. carries out intensive research, to reach the level of practical application.Towards this important medical service robot of intelligent wheel chair, carry out the exploitation based on the intelligent wheel chair navigation system of intelligent space and application study, not only can obtain the scientific and technological achievement with international most advanced level and huge market potential, effectively can also advance the work of helping the disabled of helping the elderly, serve the construction of accessible society.

Summary of the invention

In view of this, the object of the present invention is to provide a kind of hybrid locating method and system of the radio signal attenuation model based on intelligent space, by being improved the positioning precision of intelligent space intelligent wheel chair while the motion of ZigBee radio sensing network signal controlling intelligent wheel chair by the hybrid locating method of radio signal attenuation model.

For achieving the above object, the invention provides following technical scheme:

Based on a hybrid locating method for the radio signal attenuation model of intelligent space, comprise the following steps: step one: the accuracy being improved the RSSI value collected by gaussian filtering and averaging method; Step 2: use the LQI assist location algorithm that multivariable is supported, is combined by RSSI and the LQI mixed positioning algorithm after improvement and positions; Step 3: adopt weighting dynamic optimization algorithm again, further increases the precision of indoor positioning.

Present invention also offers a kind of hybrid location system of the radio signal attenuation model based on intelligent space, this system comprises CC2430/31 module, computer, intelligent wheel chair body; CC2430/31 module is used for carrying out acquisition process and the transmission of control command, is also responsible for configuration process and the transmission of RSSI value and LQI value and reference coordinate and the elements of a fix simultaneously; Computer, as the host computer of whole control system, for the point to point connect with wheelchair, completes the transmission of wheel chair sport instruction; Intelligent wheel chair is the slave computer of whole control system, is used for verifying the realization of motion control.

Further, when adopting this system to realize location, system improves the accuracy of the RSSI value collected by gaussian filtering and averaging method; And the LQI assist location algorithm using multivariable to support, combined by RSSI and the LQI mixed positioning algorithm after improvement and position; Adopt weighting dynamic optimization algorithm again, further increase the precision of indoor positioning.

Beneficial effect of the present invention is: the average range error obtained based on the mixed positioning algorithm of radio signal attenuation model provided by the invention is then at about 1 meter, greatly enhance the accuracy of range finding, the distance obtained and actual distance measurement value very close, the problem that the range error recorded under solving true environment is to a certain extent excessive, thus the positioning precision of whole navigation system is greatly improved, to meet more indoor positioning application study.

Accompanying drawing explanation

In order to make object of the present invention, technical scheme and beneficial effect clearly, the invention provides following accompanying drawing and being described:

Fig. 1 is the schematic flow sheet of the method for the invention;

Fig. 2 is the graph of a relation of RSSI and distance d;

Fig. 3 is the graph of a relation of LQI and distance d;

Fig. 4 is the range error comparison diagram of different conditions.

Embodiment

Below in conjunction with accompanying drawing, the preferred embodiments of the present invention are described in detail.

Fig. 1 is the schematic flow sheet of the method for the invention, as shown in the figure, this method improves the accuracy of the RSSI value collected by gaussian filtering and averaging method, to some extent solve the problem that range error that the randomized jitter because of RSSI causes is excessive, then certain limitation is had due to carry only more single based on the location algorithm of RSSI of hardware, in order to improve its algorithm, the LQI assist location algorithm that this method adopts multivariable to support, by RSSI and the LQI mixed positioning algorithm after improvement, add weighting dynamic optimization method, finally make to measure the range error come greatly to reduce, thus further increase the precision of indoor positioning.

Present invention also offers a kind of hybrid location system of the radio signal attenuation model based on intelligent space, this system comprises CC2430/31 module, computer, intelligent wheel chair body; CC2430/31 module is used for carrying out acquisition process and the transmission of control command, is also responsible for configuration process and the transmission of RSSI value and LQI value and reference coordinate and the elements of a fix simultaneously; Computer, as the host computer of whole control system, for the point to point connect with wheelchair, completes the transmission of wheel chair sport instruction; Intelligent wheel chair is the slave computer of whole control system, is used for verifying the realization of motion control.

In order to obtain RSSI actual value under indoor complex environment, and the relation of RSSI value and corresponding distance is studied.In the present embodiment, CC2430, CC2431 chip of employing TI company and the CC2431 navigation system solution of wireless imperial company are as experiment porch, complete the establishment of the fixer network in space, internodal data communication and RSSI signal processing and analysis, and carry out location algorithm improvement.Starting stage has carried out following experiment:

Experimental site is selected in Information barrier-free research and development centre big experiment room, the room collected using this room as RSSI value, in the selected origin of coordinates in the corner place in room, be placed in the origin of coordinates place in room with reference to node, node location is determined and in the software of location development system, is configured corresponding parameter.After configuring each parameter, start location node, collect RSSI value by location node.Method is that location node is moved in a room, the actual distance place different from this reference node is selected to test, distance is set as every one meter of record one group of RSSI value of just stopping, often organize the RSSI value of packet containing 50 corresponding test points, the frequency often organizing data acquisition is 1HZ, gathers 20 groups of data altogether.

After obtaining these data, the RSSI value collected is analyzed.In actual location environment, wireless signal is in attenuation process, and along with the change of environment, attenuation degree is also along with change.Generally speaking, RSSI value should meet certain distribution within the scope of certain, is conveniently averaging, allows measured value more restrain.Reduce random error by the improvement of algorithm again, the RSSI value finally obtained more close to actual value, thus reaches the object improving positioning precision.

Show out in fig. 2 by the description of point after the RSSI value collected is averaging by this method, their correspondence be RSSI value in different distance situation, these points are coupled together, RSSI value and signal power proportional, then also proportional with the decay of signal, therefore obtain signal attenuation curve here.The range finding theoretical model of RSSI is depicted as corresponding theoretical value curve and actual curve formation contrast more in addition by matlab.From Fig. 2, time the distance can seeing when between measured node is greater than 5 meters, the RSSI value of actual measurement is obviously greater than the RSSI value on theoretical curve.The curve decay steeper of theoretical value, and real curve is more steady.This explanation theoretical value does location algorithm, will have certain error in practical situations both.By analysis so, can see that locating wireless signal intensity is here easy to be interfered under true environment, comparatively large to the relative fluctuation of theoretical value, therefore also have larger impact to the calculating of the last elements of a fix.Herein by the RSSI value of repeatedly screening measurement, improve the accuracy of the RSSI value collected.

Same under the environment in laboratory, utilize CC2340 and CC2431 experiment porch to collect 1500 groups of RSSI value.Statistical analysis is carried out to the RSSI value collected, adopts gaussian filtering to carry out the process of RSSI value in the present embodiment.

Gauss of distribution function is as formula (1):

$f\left(x\right)=\frac{1}{\sqrt{2\mathrm{\π}}\mathrm{\σ}}{\mathrm{\σ}}^{\frac{{(x-\mathrm{\μ})}^2}{2{\mathrm{\σ}}^2}}---\left(1\right)$

In formula (1):

$\mathrm{\μ}=\frac{1}{n}\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{x}_i,{\mathrm{\σ}}^2=\frac{1}{n-1}\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{(x_i-\mathrm{\μ})}^2---\left(2\right)$

Given probable range is that 0.6 < < F (x) < < 1 draws: μ-0.85 σ≤x≤μ+0.85 σ, therefore the scope of RSSI value is after filtering: [μ-0.85 σ, μ+0.85 σ]

In result,

$\mathrm{\&sigma;}=\sqrt{\frac{1}{n-1}\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{({\mathrm{RSSI}}_i-\frac{1}{n}\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{\mathrm{RSSI}}_i)}^2}$

$\mathrm{\&mu;}=\frac{1}{n}\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{\mathrm{RSSI}}_i---\left(3\right)$

CC2431 positioning set by introducing above in laboratory, the spacing different situations of getting 6 reference nodes and a location node are tested, and record 100 groups of RSSI value at each anchor point place.Carry out filtering process by the two kinds of methods introduced above, finally compare the better off of which kind of filtering, details are as shown in table 1, can see from table 1, gaussian filtering better effects if in filtering is selected.Filtered RSSI measured value is smoother, to some extent solves the randomized jitter problem that RSSI value is affected by environment like this.

Table 1

RSSI value is stored in the nonvolatile memory of CC2431 chip, directly can obtain data from the BUF of register.And the reflection of LQI value is quality of connection, it is obtained by the structure in navigation system protocol stack.The quality that LQI value is connected with signal is directly proportional, and its value larger explanation channel quality of connection is better.In this solution of CC2431, what RSSI value and LQI value can reflect that environmental factor decays on indoor radio signal affects situation.The single distance calculated by RSSI value, then the coordinate of being tried to achieve by engine of positioning is in fact and inaccuracy, therefore this method tries to achieve the relation with distance d by collecting many group LQI values, completes the assist location to RSSI value finally by dynamic optimization.Utilize matlab simulation software, adopt least square curve fit to draw the relation of LQI value and measured node spacing d.The mode of collecting LQI value is here the same with the mode of collecting RSSI, and the place of just extracting respective value is different, and one is in the register of hardware, and one is in the packet of Software Protocol Stack.Record the distance of multiple location node to reference node equally, and corresponding LQI value, composition organizes key-value pair { (d1, LQI1) (d2, LQI2) (d3, LQI3) more ... (dn, LQIn) }.The relation curve of LQI and distance d under actual environment can be obtained, as shown in Figure 3 according to the data of these key-value pairs.

In matlab simulated environment, these LQI values obtained can be carried out 3 power curves and the 5 power curves of least square method, curve figure as shown in Figure 3.The better effects if of 5 power curves as we can see from the figure.Therefore the present embodiment adopts the algorithm of 5 power least square curve fitting formula as LQI value.Wherein the dynamic range of LQI is higher than RSSI.

LQI(d)＝p ₃d ³+p ₂d ²+p ₁d+p ₀(4)

Wherein, p ₀=172.15, p ₁=-4.6778, p ₂=0.7417, p ₃=-0.03068,

LQI(d)＝p ₅d ⁵+p ₄d ⁴+p ₃d ³+p ₂d ²+p ₁d+p ₀(5)

Wherein, p ₀=169.4996, p ₁=-1.4184, p ₂=0.0273, p ₃=-0.002, p ₄=0.000074p ₅=-0.0000048.

Because the LQI value of certain distance of polynomial solving there will be multiple situation, therefore there is code and write the inaccurate situation of value.And RSSI and LQI also has certain relation, relational expression is as shown in formula (6).

RSSI＝-(81-(LQI×91)/255) (6)

Then by evaluation fitting method, LQI value is fitted to the logarithmic curve close to 5 order polynomials, obtain the relational expression of LQI and distance d as shown in formula (7).

LQI(d ₂)＝172.3348-28.5536lg(d ₂) (7)

Through process above, codes implement can greatly reducing the computational process that euclidean distance between node pair solves.The slope of the polynomial curve of 5 matchings can also be obtained in addition by differentiate, and obtain LQI value accurately by the contrast of slope.So just can take the localization method that RSSI or LQI two kinds is different in different distances, calculate closer to the true distance value measured with most suitable algorithm, thus solve the RSSI value problem that actual measured value deviation theory curve too much causes range error excessive after actual range is more than 5 meters.

The present invention adopts two kinds of location algorithms to try to achieve the distance of location node to reference node respectively, and the parameter RSSI value of these two kinds of methods and LQI value have again certain linear relationship.Therefore the method for proposition dynamic weighting improves range finding accuracy.Distance between this laboratory environment lower node within 5 meters, the degrade condition coincidence theory model of RSSI value, therefore adopt the RSSI value distance of trying to achieve, the measured distance d between the reference node obtained by simplified model and anchor point ₁more accurate.In like manner, the distance recording point-to-point transmission by LQI is d ₂.By the dynamic optimization to gained distance, improve the accuracy of finding range between node.The method is mainly weighted process, and method of weighting is as shown in formula (8).

d＝α ₁d ₁+α ₂d ₂(8)

Consider the environment factor in laboratory and the result of previous experiments gained, a selected threshold value 5 meters.When | d ₁-d ₂|≤4m, at this moment needs to revise LQI value, is added the wobble variation value Δ LQI of self, then obtain d by revised value ₂and be brought in ranking operation above.When | d ₁-d ₂| > 4m, at this moment needs to revise RSSI value, is added the wobble variation value Δ RSSI of self, then obtain d by revised value ₁and the ranking operation be brought into above.The selection of different environmental right weight values also can have influence on positioning result.When distance when between reference node and location node that what the present embodiment adopted is is less than 5 meters, d ₁weighted value coefficient be 0.8, d ₂weighted value coefficient be 0.2, when the distance when between reference node and location node is greater than 5 meters, otherwise then.Because the method relates to the correction of shake before and after RSSI and LQI value, therefore than general dynamic weighting better effects if.

By the introduction of algorithm improvement above, the locating platform of CC2431 solution is adopted to complete the acquisition of the range error after the innovatory algorithm of 3 kinds of different modes.The range error based on RSSI algorithm respectively, based on the range error of the range error of LQI algorithm and the mixed positioning algorithm of RSSI+LQI.The average range error extracting lower 16 test points of algorithms of different after innovatory algorithm compares, and result as shown in Figure 4.

What select due to the present embodiment is attenuation model based on wireless signal, main collection be RSSI value, again because this signal value exists certain shake in same sample point, finally cause larger range error, by the analysis to sampled data, the method for gaussian filtering is adopted to carry out filtering to RSSI value.Become more mild by the RSSI value after gaussian filtering, so just can solve RSSI value excessive and add the problem of range error in test point place jitter range, thus indirectly improve positioning precision.In addition, that the present embodiment is selected is CC2431 solution party, the location algorithm that hardware system in scheme carries only has RSSI value to complete location algorithm, its locating effect is bad, and link-quality LQI and RSSI linear while also meet the attenuation relation with distance, so adopt the LQI assist location algorithm that multivariable is supported, first obtain one by curve-fitting method and compare the relation formula meeting LQI value under true environment and distance, simultaneously, dynamic weighting method is adopted to optimize the distance of measured point-to-point transmission, compensate for the only RSSI value location algorithm problem that range accuracy is inaccurate under actual range is greater than 5 meters of situations, further reduce range error, thus improve overall positioning precision.

By the contrast of the experimental data of test point in Fig. 4, can show that the average range error only using RSSI value to obtain is at about 2.5 meters, only use the average range error of LQI value acquisition at about 3.5 meters, the average range error obtained based on the mixed positioning algorithm of radio signal attenuation model in this paper is then at about 1 meter, greatly enhance the accuracy of range finding, the distance obtained and actual distance measurement value closest, the problem that the range error recorded under solving true environment is to a certain extent excessive, thus the positioning precision of whole navigation system is greatly improved, to meet more indoor positioning application study.

What finally illustrate is, above preferred embodiment is only in order to illustrate technical scheme of the present invention and unrestricted, although by above preferred embodiment to invention has been detailed description, but those skilled in the art are to be understood that, various change can be made to it in the form and details, and not depart from claims of the present invention limited range.

Claims (3)

1., based on a hybrid locating method for the radio signal attenuation model of intelligent space, it is characterized in that: comprise the following steps:

Step one: the accuracy being improved the RSSI value collected by gaussian filtering and averaging method;

Step 2: use the LQI assist location algorithm that multivariable is supported, is combined by RSSI and the LQI mixed positioning algorithm after improvement and positions;

Step 3: adopt weighting dynamic optimization algorithm again, further increases the precision of indoor positioning.

2. based on a hybrid location system for the radio signal attenuation model of intelligent space, it is characterized in that: this system comprises CC2430/31 module, computer, intelligent wheel chair body; CC2430/31 module is used for carrying out acquisition process and the transmission of control command, is also responsible for configuration process and the transmission of RSSI value and LQI value and reference coordinate and the elements of a fix simultaneously; Computer, as the host computer of whole control system, for the point to point connect with wheelchair, completes the transmission of wheel chair sport instruction; Intelligent wheel chair is the slave computer of whole control system, is used for verifying the realization of motion control.

3. based on a hybrid location system for the radio signal attenuation model of intelligent space, it is characterized in that: when adopting this system to realize location, system improves the accuracy of the RSSI value collected by gaussian filtering and averaging method; And the LQI assist location algorithm using multivariable to support, combined by RSSI and the LQI mixed positioning algorithm after improvement and position; Adopt weighting dynamic optimization algorithm again, further increase the precision of indoor positioning.