CN101592918A - Three-dimensional emergency escape guiding method based on wireless sensor network - Google Patents

Abstract

A kind of three-dimensional emergency escape guiding method based on wireless sensor network may further comprise the steps: 1) node initializing: the weights＜l of each joint is set, w 〉, wherein, l represents the hazard level of the residing floor of node; W represents the hazard level of node in this floor; And set following parameter: constant D, A _i ^I, A _iAnd e _{I, j}2) bootup process: after sensor node x sensed fire, self weights was set to A _DisAnd simultaneously the broadcasting emergency message is to network, and other nodes are received behind the urgent message channeling direction that decides self by calculating; After a sensor node y receives the urgent message that neighbor node z sends, sensor node y A _zUpgrade the weights of z in the neighbor list, node y points to the neighbours of weights minimum.The present invention can obtain best best-effort path, have good reliability.

Description

Three-dimensional emergency escape guiding method based on wireless sensor network

Technical field

The present invention relates to a kind of the building building in emergency escape guiding method.

Background technology

Along with the quickening of Chinese Urbanization's process, skyscraper, complexity are built more and more general.In the place of many public gatherings, as buildingss such as hotel, hotel, nightclub, dancing hall, cinema, small commodity markets, its inner passage is intricate, and personnel's amount of flow is huge.In a single day have disaster (as fire) to take place in these buildingss, if do not flee from effective guiding of disaster scene, consequence is with hardly imaginable.Show China's 15.9 ten thousand of breaking out of fires (not comprising forest, grassland, army, mine under ground portion fire) altogether in 2007, dead 1418 people, injured 863 people, more than 9.9 hundred million yuan of direct property loss according to the fire-fighting statistics.Wherein, be total to 8581 of breaking out of fires, dead 156 people, injured 150 people, nearly 186,790,000 yuan of direct loss as markets, hotel, public place of entertainment, hospital of school etc. at crowded place.Complexity is built, how crowded place effectively evacuates the internal staff flees from the disaster scene and organizes the external staff that rescue is arranged, and is the problem that people were concerned about always.

If can develop a kind of escape guiding system, when emergencies such as fire take place, can be according to the situation of prevailing circumstances, indicate best best-effort path quick, intelligently, make the interior of building escape personnel to flee from the disaster scene rapidly, in an orderly manner according to this indication, simultaneously can detect the fire scene again, help the external staff accurately to locate the fire place, auxiliary external rescue personnel more effectively organize rescue work, the problem of emergency escape and emergency relief difficulty in the time of so just can solving breaking out of fire in the complex building to a certain extent.The development of this emergency escape guiding system is of great immediate significance.

Traditional emergency escape indication mechanism generally all is made up of the indicator of the some directions of several static sensings, and this sensing generally all is to point to the emergency exit.But because the complicacy of buildings should be pointed to the just not necessarily sensing of a safety when fire takes place.So the sensing of control indicator makes it to point to correct escape direction, two kinds of control modes is just arranged: wired and wireless mode.Because wired control mode mainly exists two defectives: (1) has the restriction of wiring in buildings, increased cost, and if fire circuit is burnt, just will lose control to indication mechanism; (2) wired system all is the best path of navigation when adopting centralized computing method to calculate the fire generation, but there is efficiency, system will detect the position that fire takes place earlier, calculates optimal path by central control system then, goes to control the sensing of indicator again by wired mode.Be exactly in addition when fire spread, system repeats top process again once more and removes to recomputate optimal path again one time by central control system, and this will have a strong impact on the efficient of system.Based on above consideration, adopt wireless mode to have superiority more, be mainly reflected in following some: (1) does not have the restriction of wiring, and indicator can be arranged in any position in the buildings as requested, and cost also will reduce greatly; (2) total system adopts the mode of Distributed Calculation to calculate best-effort path, and whole computing velocity also will be faster than centralized computing method, and can be owing to the operation that influences very much total system of burning of certain several point.But the defective that existing wireless mode escape guiding method exists is: can not obtain best best-effort path, reliability is relatively poor.

Summary of the invention

In order to overcome the best-effort path that can not obtain the best of existing escape guiding method, the deficiency that reliability is relatively poor the invention provides and a kind ofly can obtain best best-effort path, the three-dimensional emergency escape guiding method based on wireless sensor network with good reliability.

The technical solution adopted for the present invention to solve the technical problems is:

A kind of three-dimensional emergency escape guiding method based on wireless sensor network may further comprise the steps:

1) node initializing: the weights＜l of each joint is set, w 〉, wherein, l represents the hazard level of the residing floor of node, equals the residing floor of node under initial situation; W represents the hazard level of node in this floor, and the jumping figure between described node and the adjacent node is 1;

Initial phase divides two steps to finish: the first, calculate each node to the outlet of his place floor or the minimum hop count of stair node; The second, utilize node to come the initial weight of computing node to the minimum hop count of outlet or stair node.

(1.1) calculate each node to the outlet of its place floor or the minimum hop count of stair node

This process divides floor to carry out respectively.This process is broadcasted INIT start of message (SOM) by Egress node.The form of INIT message is as follows:＜e_id, f_id, hop 〉.Three fields represent to export node i d, forward node id and the forward node jumping figure to Egress node respectively.One Egress node x at first broadcasts INIT message (x, x, 0), and node y judges if node x is new Egress node then is added into the outlet tabulation after receiving this message, and notes jumping figure h _xIf not new Egress node and h _x＞hop+1, then h _x=hop+1, reconstruct INIT message (x, y, h then _x) and transmit.Last each node calculates apart from the jumping figure of self nearest Egress node and is recorded in h _cIn.Other floors can be regarded the stair node as the Egress node of this layer, and processing procedure is identical.

(1.2) calculating of node initial weight

The account form of node initial weight is as follows:

$w_i=\left\{\begin{array}{cc}{w}_{\mathrm{dis}}\×\mathrm{\α}\×h_c\×e^{-\underset{j=1}{\overset{m}{\mathrm{\Σ}}}\frac{1}{h_j}},& {\mathrm{node}}_i\∈\mathrm{Normal}\\ 0,& {\mathrm{node}}_i\∈\mathrm{Exit\; or\; Stair}\end{array}\right.---\left(1\right)$

w _DisRepresent the maximum weights that node can arrive, this value will be used to give the weights of the node that disaster takes place.M is the number of this layer outlet or stair, and algorithm uses

Weigh the hop of each Egress node of nodal distance, and get each hop reciprocal and negative as index, be referred to as the comprehensive distance index.h _cSpan is from the jumping figure of the nearest Egress node of this node.Parameter alpha is used to weigh initial weight and accounts for w _DisProportion.

The initial floor weights of node are set to the residing floor of node.

Each node can be preserved a form that comprises information of neighbor nodes, comprising numbering, type, weights and the position of neighbor node.

2) bootup process:

Set following parameter:

D a: constant.If the distance of certain sensor and any disaster origination point all is less than or equal to this constant, then claim this sensor to be positioned at the hazardous location;

A _Dis: the weights (l that detects the sensor of fire _Dis, w _Dis);

A _i ^I: the initial weight (l of sensor i _i ^I, w _i ^I);

A _i: the weights (l of sensor i _i, w _i);

e _{I, j}: from a sensor i who detects emergency to the jumping figure another sensor j.

After sensor node x sensed fire, self weights was set to A _DisAnd broadcasting emergency message EMG simultaneously, other nodes are received behind the urgent message channeling direction that decides self by calculating; After a sensor node y received the urgent message that neighbor node z sends, sensor node y at first used A _zUpgrade the weights of z in the neighbor list, node y carries out according to the following steps then:

(2.1) node y judges whether x is new disaster information, comes the jumping figure e to the disaster node of new node more with this _{X, y}:

(2.1.1) if new disaster message, node y is to the jumping figure e of disaster node x _{X, y}Be updated to (hop+1);

(2.1.2) if not new disaster message, node y judges that whether hop+1 is less than e _{X, y}, if e then _{X, y}Be updated to (hop+1); Otherwise do not make an amendment;

(2.2) node y judges e _{X, y}Whether be less than or equal to D, if set up then represent that node y is positioned within the hazardous location, will raise immediately self weights of node y are represented; If the lower floor section of ladder point of a stair node is in the hazardous location, think that then these stair also are in the hazardous location, self weights also will raise;

If (2.2.1) y is an ordinary node, and e _{X, y}≤ D then is provided with:

l _y＝l _emg；

$w_y=\mathrm{MAX}\{w_y,w_y^I+w_{\mathrm{dis}}\×{\mathrm{\βe}}^{-\mathrm{\μ}{h}_d}\},$ Wherein $\mathrm{\β}=\frac{h_c+1}{h_d+1}---\left(2\right)$

The increment of node weights is weighed with the jumping figure of the nearest disaster node of nodal point separation.β is a update coefficients, is used to reflect gathering way of weights, h _cWith h _dThe minor increment of representing nearest Egress node of present node and nearest disaster node respectively.The weights that parameter μ is used to limit after the renewal exceed maximum weight w _DisAnd consideration w _y ^lValue be because, same point in the hazardous location still can be different from the distance of nearest outlet, hazard level also can be different.Use max to obtain the maximum weights that in a plurality of disasters, can adjust simultaneously.

If (2.2.2) y is the stair nodes, then be provided with by following rule:

If 1. $l_z^I<l_y^I$ And e _{X, y}-1≤D, then node y thinks that self is in the hazardous location, because its lower level node is in the hazardous location.E is set this moment _{X, y}=hop, l _y=l _Emg, w _y=w _z-ε.Wherein ε is a constant, and this constant depends on the maximum node number of a floor.

If 2. $l_z^I\&GreaterEqual;l_y^I$ And e _{X, y}≤ D thinks that then node y is in the hazardous location, is provided with l _y=l _Emg, w _yBe provided with according to formula (2);

(2.3) node y judges self whether to become locally minimum, and promptly node y self is not an Egress node, but its weights are but less than its weights of all neighbor nodes; Judgment rule is as follows: if y is an ordinary node, then to its all neighbours x, w is arranged _y＜w _xIf y is the stair nodes, then, A is arranged to its all neighbours x _y＜A _x

If (2.3.1) node y is an ordinary node, then y adjusts the weights of self:

Be all neighbours' weights; Avg () is the mean value of all neighbours' weights; Max () is the maximal value of neighbours' weights; δ is the constant of a fractional value, and this constant guarantees convergence of algorithm; λ is the constant of a fractional value; C is that node becomes local minimum number of times continuously; K is that node becomes the local minimum maximum times that is allowed continuously.

If (2.3.2) node y is the stair nodes, then node y is according to l _yAdjust its weights, stair node guiding personnel specifically have toward downstairs, upstairs or with other stair nodes of layer evacuate:

If 1. $l_y=l_y^I,$ Y is provided with l _y=l _Emg, w _yAdjust according to formula (3).At this moment the lower level node of y is in the hazardous location, so y should guide personnel toward other stair node guiding with layer.

If 2. l _y=l _Emg, y is provided with l _y=l _Dis, at this moment y judge it the l value of upper layer node whether less than l _Dis, if y is provided with w _y=0, at this moment because the lower floor section of ladder point of y is in the hazardous location, and also covered, so y can guide up floor ladder evacuation of personnel by the everybody towards road with other stair of floor.Otherwise the w of node y _yAdjust according to formula (3).

If 3. l _y=l _Dis, at this moment y keeps l _yConstant, w _yAdjust according to formula (3).At this moment y can seek a best-effort path that hazard level is relatively little.

(2.4) if node y meets the following conditions, then continue the broadcasting emergency message:

If (2.4.1) node y receives it is a new urgent message;

If (2.4.2) e takes place to the jumping figure of disaster node x in node y _{X, y}Change;

(2.4.3) the weights A of node y _yChange;

(2.5) node y points to the neighbours of weights minimum:

If (2.5.1) y is an ordinary node, then node y points to the neighbours of weight w minimum among the neighbours;

If (2.5.2) y is the stair nodes, then node y points to weights A (l, w) Zui Xiao neighbours among the neighbours.

As preferred a kind of scheme: in described node bootup process, EMG is set promptly to be wrapped, include five fields in the urgent bag of EMG: (1) finds the sensor ID of fire, (2) ID of forward node, (3) jumping figure from forward node to the sensor node that detects fire, (4) weights of forward node, the type of (5) forward node.

Node weights A in the urgent bag of EMG _iComprise two field (l _i, w _i), l wherein _iThe hazard level of expression node floor of living in, w _iThe hazard level of representing self.In node is in the hazardous location time, composing the floor weights to node is l in algorithm _Emg

Further, described wireless sensor network is the Zigbee network.

Technical conceive of the present invention is: (Wireless Sensor Network, WSN) technology is that the semiconductor technology, wireless communication technology in 20th century and a new information that the development of MEMS (micro electro mechanical system) integrated high-speed is bred are obtained and processing mode to wireless sensor network.WSN is a kind of network that is made of a series of sensor nodes, be used for monitoring in real time, the various information (as physical phenomenons such as light intensity, temperature, humidity, noise and harmful gas concentrations) of the interested perceptive object of observer in perception and the acquisition node deployment region, and send wirelessly after these information are handled, finally send to the supervisor by wireless network.

The ZigBee agreement is based on IEEE 802.15.4 standard and the network protocol standard formulated.ZigBee is the standard wireless network protocol stack that aims at low rate sensor and Control Network design, is best suited for the standard of wireless sensor network.It is a kind of short distance, low rate, low-power consumption, low cost and low complex degree.Compare with other wireless protocols, the ZigBee agreement has following characteristics: the data rate rate is low, low in energy consumption, cost is low, network capacity is big, time delay is short, the self-organization of network, self-healing ability strong, reliable communications.ZigBee works in the 2.4GHz ISM band that need not to register, and transfer rate is 250kb/s, and transmission range is 10～75m.It has battery life long, use characteristics such as simple, that reliability is high and networking capability is strong.

Utilize a kind ofly, realize having the three-dimensional building emergency escape guiding system of indoor security monitoring based on the Zigbee network technology.At ordinary times, wireless sensor network node monitoring buildings has or not anomalous event to take place in the system, for example incident such as fire.When having emergency to take place in the scope of being monitored, sensor node will start a kind of intelligent algorithm, it will calculate the best direction indication of each node automatically, when disaster spreads variation, the place of former safety also might become dangerous even become the place that new disaster takes place, algorithm will help the escape crowd to flee from the disaster scene along with the direction indication of each node of renewal that spreads change dynamics of disaster when the time comes.

In the buildings of a three-dimensional structure, with the wireless sensing node of rational layout placement some, form wireless sensor network, environmental parameters such as the temperature around these node monitorings, smog take place to have judged whether fire.In case the environmental data identification that comes according to collection when certain or some nodes has fire to take place, these nodes will enter abnormality, and broadcast this abnormality to whole network.The node of receiving abnormality is according to information such as fire occurrence positions, self-position, exit positions, calculate a safe escape direction in real time, and using Warning Mark that this direction is pointed out, the trapped personnel in the buildings can be escaped according to these Way outs to nearby that points to step by step.

The present invention weighs the hazard level of each node with the notion of weights, high more this node of expression of weights zone of living in is dangerous more, the guiding in path will progressively be pointed to the lower some guiding of weights from the higher point of weights, until guiding to nearest Egress node, and represent euclidean distance between node pair with the minimum hop count between two nodes.In disaster take place period, all node weights all will dynamically update, and arrive safest outlet with the guiding personnel.

Beneficial effect of the present invention mainly shows: when disaster takes place, can calculate correct channeling direction rapidly guides the escape crowd to flee from the disaster scene, here emphasis is meant the three-dimensional building escape, variation intelligence according to the condition of a disaster changes channeling direction simultaneously, thereby avoids the escape crowd blindly to escape.When avoiding current danger, avoid taking a roundabout way, in hazard level and escape time, average out; Avoided occurring guiding loop situation; And modern architecture often has a more than fire exit, also obtained balance in escape time and outlet are selected.

Description of drawings

Fig. 1 is the layout of node in the three-dimensional building thing among the present invention

Fig. 2 is the process flow diagram of three-dimensional building emergency escape guiding method.

Embodiment

Below in conjunction with accompanying drawing the present invention is further described.

With reference to Fig. 1～Fig. 2, a kind of three-dimensional building emergency escape guiding method based on wireless sensor network, this method is divided into two parts: (1) initial phase; (2) vectoring phase.

(1) initial phase

The purpose of initial phase is to the safe guidance of egress under regular situation, is exactly the safe guidance direction of conventional time as the arrow among the figure one.In the 3D environment, represent the hazard level of node with weights, weights are low more safe more, and the weights minimum of Egress node is zero, and the weights of fire node be maximum.Because 3D environment lower node is in different floors, so the weights of composing for each node are made up of two fields, form is＜l w 〉, wherein l represents the hazard level of the residing floor of node, equals the residing floor of node under initial situation; W represents the hazard level of node in this floor.When calculating channeling direction, we are that the height according to the node weights comes the high node of guiding personnel value as a matter of expediency to evacuate to the low node of weights.So when the weights of comparison node, if ordinary node then can be ignored first field l, and only compare second field w, to point to the neighbor node of same floor; If the stair node then first field also to compare, be to point to determine this stair node with the neighbor node of layer or stair node upstairs or downstairs.Therefore, the magnitude relationship of defined node weights: if to ordinary node w _x＜w _y, weight then _x＞weight _yIf to stair node (l _x＞l _y) or (l _x=l _yAnd w _x＞w _y), weight then _x＞weight _y

Initial phase divides two steps to finish: the first, calculate each node to the outlet of his place floor or the minimum hop count of stair node; The second, utilize node to come the initial weight of computing node to the minimum hop count of outlet or stair node.

1. calculate each node to the outlet of its place floor or the minimum hop count of stair node

This process divides floor to carry out respectively.Be that example illustrates the computation process of node to the jumping figure of Egress node with the floor that comprises Egress node now, other floors can be regarded the stair node as the Egress node of this layer, and computing method are identical.This process is broadcasted INIT start of message (SOM) by Egress node.The form of INIT message is as follows:＜e_id, f_id, hop 〉.Three fields represent to export node i d, forward node id and the forward node jumping figure to Egress node respectively.One Egress node x at first broadcasts INIT message (x, x, 0), and node y judges if node x is new Egress node then is added into the outlet tabulation after receiving this message, and notes jumping figure h _xIf not new Egress node and h _x＞hop+1, then h _x=hop+1, reconstruct INIT message (x, y, h then _x) and transmit.Last each node calculates apart from the jumping figure of self nearest Egress node and is recorded in h _cIn.The processing procedure of other floors is identical.

2. the calculating of node initial weight

The account form of node initial weight is as follows:

$w_i=\left\{\begin{array}{cc}{w}_{\mathrm{dis}}\&times;\mathrm{\&alpha;}\&times;h_c\&times;e^{-\underset{j=1}{\overset{m}{\mathrm{\&Sigma;}}}\frac{1}{h_j}},& {\mathrm{node}}_i\&Element;\mathrm{Normal}\\ 0,& {\mathrm{node}}_i\&Element;\mathrm{Exit\; or\; Stair}\end{array}\right.---\left(1\right)$

w _DisRepresent the maximum weights that node can arrive, this value will be used to give the weights of the node that disaster takes place.M is the number of this layer outlet or stair, and algorithm uses

Weigh the hop of each Egress node of nodal distance, and get each hop reciprocal and negative as index, we are referred to as the comprehensive distance index.Characteristic by exponential function shows that the jumping figure of nodal distance Egress node is more little, and then the initial weight of node will be more little; The node initial weight also will reduce along with increasing of Egress node simultaneously.h _cSpan is from the jumping figure of the nearest Egress node of this node, even algorithm is considered the reason of this parameter and is that the Egress node comprehensive distance index of node is identical that the less obvious bigger node of node of the jumping figure of nearest Egress node has better security.Parameter alpha is used to weigh initial weight and accounts for w _DisProportion.If node type is outlet or stair node, then its initial weight is zero.

The initial floor weights of node are set to the residing floor of node.

After finishing initial phase work, the initial weight of node and initial directional are shown in figure one.Each node can be preserved a form that comprises information of neighbor nodes, comprising numbering, type, weights and the position of neighbor node.

(2) vectoring phase

When node detected the fire generation, algorithm entered the vectoring phase.Be the several basic conceptions of using in the algorithm below:

D a: constant.If the distance of certain sensor and any disaster origination point all is less than or equal to this constant, then claim this sensor to be positioned at the hazardous location;

A _Dis: the weights (l that detects the sensor of fire _Dis, w _Dis);

A _i ^I: the initial weight (l of sensor i _i ^I, w _i ^I);

A _i: the weights (l of sensor i _i, w _i);

e _{I, j}: from a sensor i who detects emergency to the jumping figure another sensor j.

EMG promptly wraps: include five fields in the urgent bag of EMG: (1) finds the sensor ID of fire, (2) ID of forward node, (3) jumping figure from forward node to the sensor node that detects fire, the weights of (4) forward node, the type of (5) forward node.The urgent bag of EMG can return the data of emergency, and helps sensor to adjust their weights, to reach the purpose of guiding.

Node weights A in the urgent bag of EMG _iComprise two field (l _i, w _i), l wherein _iThe hazard level of expression node floor of living in, w _iThe hazard level of representing self.In node is in the hazardous location time, composing the floor hazard level to node is l in algorithm _Emg, therefore work as the stair node on upper strata and find that the floor hazard level of lower level node is more than or equal to l _EmgThe time, can avoid personnel are directed to its lower level node.

After sensor node x sensed fire, self weights was set to A _Dis, and broadcast the urgent message of EMG (x, x, an A simultaneously _Dis, 0, type).This urgent message will be addressed in the network, and other nodes are received behind the urgent message channeling direction that decides self by calculating.When a sensor node y receives urgent message EMG (x, z, the A that neighbor node z sends _z, hop, type) after, y at first uses A _zUpgrade the weights of z in the neighbor list, node y carries out according to the following steps then:

1. node y judges whether x is new disaster information, comes the jumping figure e to the disaster node of new node more with this _{X, y}:

(1) if new disaster message, node y is to the jumping figure e of disaster node x _{X, y}Be updated to (hop+1);

(2) if not new disaster message, node y judges that whether hop+1 is less than e _{X, y}, if e then _{X, y}Be updated to (hop+1); Otherwise do not make an amendment.

If z is the stair nodes, y be ordinary node then y l is set _y=l _z, to prevent that when the step 5 y and z point to mutually because there is w _y＞w _zAnd l _y＜l _z

2. node y judges e _{X, y}Whether be less than or equal to D, if set up then represent that node y is positioned within the hazardous location, will raise immediately self weights of node y are represented.If the lower floor section of ladder point of a stair node is in the hazardous location, think that then these stair also are in the hazardous location, self weights also will raise.

(1) if y is an ordinary node, and e _{X, y}≤ D then is provided with:

l _y＝l _emg；

$w_y=\mathrm{MAX}\{w_y,w_y^I+w_{\mathrm{dis}}\&times;{\mathrm{\&beta;e}}^{-\mathrm{\&mu;}{h}_d}\},$ Wherein $\mathrm{\&beta;}=\frac{h_c+1}{h_d+1}---\left(2\right)$

The increment of node weights is weighed with the jumping figure of the nearest disaster node of nodal point separation.β is a update coefficients, is used to reflect gathering way of weights, h _cWith h _dThe minor increment of representing nearest Egress node of present node and nearest disaster node respectively.The weights that parameter μ is used to limit after the renewal exceed maximum weight w _DisAnd consideration w _y ^IValue be because, same point in the hazardous location still can be different from the distance of nearest outlet, hazard level also can be different.Use max to obtain the maximum weights that in a plurality of disasters, can adjust simultaneously.

(2), then be provided with by following rule if y is the stair nodes:

If 1. $l_z^I<l_y^I$ And e _{X, y}-1≤D, then node y thinks that self is in the hazardous location, because its lower level node is in the hazardous location.E is set this moment _{X, y}=hop, l _y=l _Emg, w _y=w _z-ε.Wherein ε is a constant, and this constant depends on the maximum node number of a floor.

If 2. $l_z^I\&GreaterEqual;l_y^I$ And e _{X, y}≤ D thinks that then node y is in the hazardous location, is provided with l _y=l _Emg, w _yBe provided with according to formula (2);

3.y whether self become local minimum, that is to say that node y self is not an Egress node if judging, but its weights are but less than its weights of all neighbor nodes.Judgment rule is as follows: if y is an ordinary node, then to its all neighbours x, w is arranged _y＜w _xIf y is the stair nodes, then, A is arranged to its all neighbours x _y＜A _x

(1) if y is an ordinary node, then y adjusts the weights of self:

All neighbours' weights.

Avg (): the mean value of all neighbours' weights.

Max (): the maximal value of neighbours' weights.

δ: the constant of a fractional value.This constant guarantees convergence of algorithm.

λ: the constant of a fractional value.

C: node becomes local minimum number of times continuously.

K: node becomes the local minimum maximum times that is allowed continuously.

The implication of formula is as follows: when node becomes local minimum number of times continuously less than k, remove the weights that raise with the method for mean value; When node becomes local minimum number of times continuously more than or equal to k, show that this node is in the dangerous situation of surrounding, the weights that should raise more apace are so deduct the next quick rising weights of method of a little constant with maximum neighbours' weights.

(2) if y is the stair nodes, then y is according to l _yAdjust its weights.The stair node can guide personnel toward downstairs, evacuate upstairs or with other stair nodes of layer.Generally speaking, the stair node is with the past guiding downstairs of personnel; But when lower floor section of ladder point was in the hazardous location, the stair node can be with personnel toward other stair node guiding with layer; If at this moment the road that leads to other stair of floor is also covered by fire, just personnel should be guided to seek other safe best-effort path to the upper strata.

If 1. $l_y=l_y^I,$ Y is provided with l _y=l _Emg, w _yAdjust according to formula (3).At this moment the lower level node of y is in the hazardous location, so y should guide personnel toward other stair node guiding with layer.

If 2. l _y=l _Emg, y is provided with l _y=l _Dis, at this moment y judge it the l value of upper layer node whether less than l _Dis, if y is provided with w _y=0, at this moment because the lower floor section of ladder point of y is in the hazardous location, and also covered, so y can guide up floor ladder evacuation of personnel by the everybody towards road with other stair of floor.Otherwise the w of node y _yAdjust according to formula (3).

If 3. l _y=l _Dis, at this moment y keeps l _yConstant, w _yAdjust according to formula (3).At this moment y can seek a best-effort path that hazard level is relatively little.

4., then continue broadcasting emergency message EMG (x, y, A if y meets the following conditions _y, e _{X, y}, type):

(1) be a new urgent message if node y receives;

(2) if node y arrives the jumping figure of disaster node x e takes place _{X, y}Change;

(3) the weights A of node y _yChange.

5. guiding.Node y points to the neighbours of weights minimum:

(1) if y is an ordinary node, then node y points to the neighbours of weight w minimum among the neighbours;

(2) if y is the stair nodes, then node y points to weights A (l, w) Zui Xiao neighbours among the neighbours.

Claims (3)

1, a kind of three-dimensional emergency escape guiding method based on wireless sensor network is characterized in that: adopt sensor network to monitor building building environment, when the generation emergency need be escaped in the building, system can provide three-dimensional emergency escape guiding method.May further comprise the steps:

1) node initializing: the height value＜l of each joint is set, w 〉, wherein, l represents the hazard level of the residing floor of node, equals the residing floor of node under initial situation; W represents the hazard level of node in this floor, and the jumping figure between described node and the adjacent node is 1;

Initial phase divides two steps to finish:

(1.1) calculate each node to the outlet of its place floor or the minimum hop count of stair node

One Egress node x at first broadcasts INIT message (x, x, 0), after node y receives the INIT message, then is added into the outlet tabulation if decision node x is new Egress node, and notes jumping figure h _xIf not new Egress node and h _x＞hop+1, then h _x=hop+1, reconstruct INIT message (x, y, h then _x) and transmit, each node calculates apart from the jumping figure of self nearest Egress node and is recorded in h at last _cIn; Can be for other floors with the Egress node of stair node as this layer, processing procedure is same as described above.

(1.2) calculating of node initial weight

The account form of node initial weight is as follows:

$w_i=\left\{\begin{array}{cc}{w}_{\mathrm{dis}}\&times;\mathrm{\&alpha;}\&times;h_c\&times;e^{-\underset{j=1}{\overset{m}{\mathrm{\&Sigma;}}}\frac{1}{h_j}},& {\mathrm{node}}_i\&Element;\mathrm{Normal}\\ 0,& {\mathrm{node}}_i\&Element;\mathrm{ExitorStair}\end{array}\right.---\left(1\right)$

w _DisRepresent the maximum weights that node can arrive, this value will be used to give the weights of the node that disaster takes place, and m is the number of this layer outlet or stair, and algorithm uses Weigh the hop of each Egress node of nodal distance, and get each hop reciprocal and negative as index, be referred to as the comprehensive distance index, h _cSpan is from the jumping figure of the nearest Egress node of this node; Parameter alpha is represented to weigh initial weight and is accounted for w _DisProportion;

The initial floor weights of each node are set to the residing floor of node;

Each node can be preserved a form that comprises information of neighbor nodes, comprising numbering, type, weights and the position of neighbor node;

2) bootup process:

Set following parameter:

Constant D:, then claim this sensor to be positioned at the hazardous location if the distance of certain sensor and any disaster origination point all is less than or equal to this constant;

A _Max: the height value (l that detects the sensor of emergency _Max, w _Max);

A _i ^I: the elemental height value (l of sensor i _i ^I, w _i ^I);

A _i: the height value (l of sensor i _i, w _i);

e _{I, j}: from a sensor i who detects emergency to the jumping figure another sensor j;

After sensor node x sensed fire, the oneself height value was set to A _MaxAnd the broadcasting emergency message will be addressed in the network simultaneously; After a sensor node y received the urgent message that neighbor node z sends, sensor node y at first used A _zUpgrade the height value of z in the neighbor list, node y carries out according to the following steps then:

(2.1) node y judges that whether x is new disaster information, specifically has:

(2.1.1) if new disaster message, node y is to the jumping figure e of disaster node x _{X, y}Be updated to (hop+1);

(2.1.2) if not new disaster message, node y judges that whether hop+1 is less than e _{X, y}, if e then _{X, y}Be updated to (hop+1); Otherwise do not make an amendment;

(2.2) node y judges e _{X, y}Whether be less than or equal to D, if set up then represent that node y is positioned within the hazardous location, raise immediately self weights of node y are represented;

If (2.2.1) y is an ordinary node, and e _{X, y}≤ D is provided with according to formula (2):

l _y＝l _emg；

$w_y=\mathrm{MAX}\{w_y,w_y^I+w_{\mathrm{dis}}\&times;{\mathrm{\&beta;e}}^{-\mathrm{\&mu;}{h}_d}\},$ Wherein $\mathrm{\&beta;}=\frac{h_c+1}{h_d+1}---\left(2\right)$

Wherein, β is a update coefficients, is used to reflect gathering way of weights, h _cWith h _dRepresent the nearest Egress node of present node and the nearest minor increment of disaster node respectively, the weights that parameter μ is used to limit after the renewal exceed maximum weight w _Dis, w _y ^IThe same point in the hazardous location of expression, from the distance of nearest outlet and the related coefficient of hazard level existence, MAX represents to obtain the maximum weights that can adjust in a plurality of disasters;

If (2.2.2) y is the stair nodes, be provided with by following rule:

If 1. $l_z^I<l_y^I$ And e _{X, y}-1≤D, then node y thinks and self is in the hazardous location, and e is set _{X, y}=hop, l _y=l _Emg, w _y=w _z-ε, wherein ε is a constant, this constant depends on the maximum node number of a floor;

If 2. $l_z^I\&GreaterEqual;l_y^I$ And e _{X, y}≤ D thinks that then node y is in the hazardous location, is provided with l _y=l _Emg, w _yBe provided with according to formula (2);

(2.3) node y judge self whether become local minimum, but its weights are but less than its weights of all neighbor nodes; Judgment rule is as follows: if y is an ordinary node, then to its all neighbours x, w is arranged _y＜w _xIf y is the stair nodes, then, A is arranged to its all neighbours x _y＜A _x

If (2.3.1) node y is an ordinary node, then y adjusts the weights of self:

Be all neighbours' weights; Avg () is the mean value of all neighbours' weights; Max () is the maximal value of neighbours' weights; δ is the constant of a fractional value, and this constant guarantees convergence of algorithm; λ is the constant of a fractional value; C is that node becomes local minimum number of times continuously; K is that node becomes the local minimum maximum times that is allowed continuously;

If (2.3.2) node y is the stair nodes, then node y is according to l _yAdjust its weights, specifically have:

If 1. $l_y=l_y^I,$ Y is provided with l _y=l _Emg, w _yAdjust according to formula (3);

If 2. l _y=l _Emg, y is provided with l _y=l _Dis, node y judges that whether its l value of upper layer node is less than l _Dis, if node y is provided with w _y=0; Otherwise the w of node y _yAdjust according to formula (3);

If 3. l _y=l _Dis, node y keeps l _yConstant, w _yAdjust according to formula (3);

(2.4) if node y meets the following conditions, then continue the broadcasting emergency message:

If (2.4.1) node y receives it is a new urgent message;

If (2.4.2) e takes place to the jumping figure of disaster node x in node y _{X, y}Change;

(2.4.3) the weights A of node y _yChange;

(2.5) node y points to the neighbours of weights minimum:

If (2.5.1) y is an ordinary node, then node y points to the neighbours of weight w minimum among the neighbours;

If (2.5.2) y is the stair nodes, then node y points to weights A (l, w) Zui Xiao neighbours among the neighbours.

2, the three-dimensional building emergency escape guiding method based on wireless sensor network as claimed in claim 1, it is characterized in that: in described node bootup process, EMG is set promptly to be wrapped, include five fields in the urgent bag of EMG: (1) finds the sensor ID of fire, (2) ID of forward node, (3) jumping figure from forward node to the sensor node that detects fire, the weights of (4) forward node, the type of (5) forward node.

Node weights A in the urgent bag of EMG _iComprise two field (l _i, w _i), l wherein _iThe hazard level of expression node floor of living in, w _iThe hazard level of representing self.In node is in the hazardous location time, composing the floor weights to node is l in algorithm _Emg

3, the three-dimensional emergency escape guiding method based on wireless sensor network as claimed in claim 1 or 2 is characterized in that: described wireless sensor network is the Zigbee network.

Images