CN101592918B - Three-dimensional emergency escape guiding method based on wireless sensor network - Google Patents

Abstract

The invention relates to a three-dimensional emergency escape guiding method based on a wireless sensor network, which comprises the following steps: 1) node initiation, namely setting a weight <1, w> of each node, wherein 1 refers to the danger level of a floor where the node is positioned, and w refers to the danger level of the node in the floor, and setting the following parameters: a constant D, Ai<I>, Ai and e[i, j]; and 2) guiding process, wherein when a sensor node x senses fire, the sensor node x sets the self weight as Adis and simultaneously broadcasts an emergency message to the network, and other nodes decide self guiding directions through calculation after receiving the emergency message; and when a sensor node y receives an emergency message sent by a neighbor node z, the sensor node y updates a weight of z in a neighbor list by Az, and directs to a neighbor with the minimum weight. The method can obtain the optimal escape path, and has 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 Urbanization Process In China, skyscraper, complex building are 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 occur 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 at crowded place, be total to 8581 of breaking out of fires as markets, hotel, public place of entertainment, hospital of school etc., dead 156 people, injured 150 people, nearly 186,790,000 yuan of direct loss.Complex building, how crowded place effectively evacuates the internal staff flees from the disaster scene and organize the external staff that rescue is arranged, and is the problem that people are concerned about always.

If can develop a kind of escaping guidance system, when the emergencies such as fire occur, 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 rapidly, in an orderly manner the disaster scene 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, emergency escape and emergency relief hard problem in the time of so just can solving to a certain extent in complex building breaking out of fire.The development of this emergency escape guiding system is of great immediate significance.

Traditional emergency escape indication mechanism generally all is comprised of the indicator of the some directions of several static sensings, and this sensing is all generally to point to the emergency exit.But due to the complicacy of buildings, when occuring, fire should point to the just not necessarily sensing of a safety.So the sensing of Control director makes it to point to correct escape direction, two kinds of control modes are 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 the control to indication mechanism; (2) wired system is all the best path of navigation when adopting the centralized computing method to calculate the fire generation, but there is efficiency, system first will detect the position that fire occurs, then calculate optimal path by central control system, then go the sensing of Control director by wired mode.Be exactly in addition when fire spread, system again repeats again top process and removes to recomputate optimal path one time by central control system again, and this will have a strong impact on the efficient of system.Based on above consideration, adopt wireless mode more to have superiority, be mainly reflected in following some: (1) does not have the restriction of wiring, and indicator can be arranged in any position in buildings as requested, and cost also will reduce greatly; (2) whole system adopts the mode of Distributed Calculation to calculate best-effort path, and whole computing velocity also will be faster than the centralized computing method, and can be due to the operation that affects very much whole 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 comprises the following 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 adjacent node is 1;

Initial phase divides two steps to complete: 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 respectively to export node i d, forward node id and forward node to the jumping figure of Egress node.At first one Egress node x broadcasts INIT message (x, x, 0), if node y judges that node x is that new Egress node is added into the outlet list after receiving this message, and records jumping figure h _xIf not new Egress node and h _x＞hop+1, h _x=hop+1, then reconstruct INIT message (x, y, h _x) and forward.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:

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

w _disRepresent the maximum weights that node can arrive, this value will be used to the weights of giving the node that disaster occurs.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 for weighing 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, 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 and are decided the channeling direction of self after urgent message by calculating; After a sensor node y received the urgent message that neighbor node z sends, at first sensor node y used A _zUpgrade the weights of z in neighbor list, then node y carries out according to the following steps:

(2.1) node y judges whether x is new disaster information, comes new node more to the jumping figure e of disaster node 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 it is e _{X, y}Be updated to (hop+1); Otherwise do not make an amendment;

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

If (2.2.1) y is ordinary node, and e _{X, y}≤ D arranges:

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 for gathering way of reflection weights, h _cWith h _dThe minor increment that represents respectively the nearest Egress node of present node and nearest disaster node.The weights that parameter μ is used for after restriction is upgraded 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.Obtain with max the maximum weights that can adjust in a plurality of difficulties of going out simultaneously.

If (2.2.2) y is the stair nodes, arrange by following rule:

If 1. $l_z^I<l_y^I$ And e _{X, y}-1≤D, 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 node y is in the hazardous location, arranges l _y=l _emg, w _yArrange according to formula (2);

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

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

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 allows continuously.

If (2.3.2) node y is the stair nodes, 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 arranges 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 arranges l _y=l _dis, at this moment y judge it the l value of upper layer node whether less than l _dis, if y arranges w _y=0, at this moment because the lower floor section of ladder point of y is in the hazardous location, and also covered by the everybody towards the road with other stair of floor, so y can guide up floor ladder evacuation of personnel.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, continue the broadcasting emergency message:

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

If (2.4.2) e occurs 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 ordinary node, node y points to the neighbours of weight w minimum in neighbours;

If (2.5.2) y is the stair nodes, node y points to the minimum neighbours of weights A (l, w) in 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 fire detected, (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 that represents self.When Nodes was in the hazardous location, composing the floor weights to node was l in algorithm _emg

Further, described wireless sensor network is Zigbee network.

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

Zigbee protocol 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 the 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, Zigbee protocol has following characteristics: the data rate rate is low, low in energy consumption, cost is low, network capacity is large, 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 the characteristics such as simple, that reliability is high and networking capability is strong.

Utilize a kind ofly based on the Zigbee network technology, realize having the three-dimensional building emergency escape guiding system of indoor security monitoring.At ordinary times, in system, wireless sensor network node monitoring buildings has or not anomalous event to occur, such as events such as fire.When having emergency to occur in the scope of monitoring, 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 the dangerous place that new disaster occurs that even becomes, algorithm will change the dynamic direction that upgrades each node along with spreading of disaster when the time comes, help the escape crowd to flee from the disaster scene.

In the buildings of a three-dimensional structure, with the wireless sensing node of rational layout placement some, form wireless sensor network, the environmental parameters such as the temperature around these node monitorings, smog occur to have judged whether fire.In case the environmental data identification that comes according to collection when certain or some nodes has fire to occur, these nodes will enter abnormality, and broadcast this abnormality to whole network.Receive that the node of abnormality is according to information such as fire occurrence positions, self-position, exit positions, calculate in real time a safe escape direction, and using Warning Mark that this direction is pointed out, the trapped personnel in 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 concept of weights, this node of the higher expression of weights zone of living in is more dangerous, the guiding in path will progressively be pointed to the lower some guiding of weights from the higher point of weights, until guide to nearest Egress node, and represent euclidean distance between node pair with the minimum hop count between two nodes.In disaster, occur period, all node weights all will dynamically update, and arrive safest outlet with the guiding personnel.

Beneficial effect of the present invention is mainly manifested in: when disaster occurs, can calculate rapidly correct channeling direction guides the escape crowd to flee from the disaster scene, here emphasis refers to 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, average out in hazard level and escape time; Avoided occurring guiding the loop situation; And modern architecture often has a more than fire exit, also obtained balance in escape time and egress selection.

Description of drawings

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

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

Embodiment

The invention will be further described below in conjunction with accompanying drawing.

With reference to Fig. 1～Fig. 2, a kind of three-dimensional building emergency escape guiding method based on wireless sensor network, the 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 in figure one.In the 3D environment, represent the hazard level of node with weights, weights are more low safer, and the weights minimum of Egress node is zero, and the weights of fire node are maximum.Because 3D environment lower node is in different floors, so the weights of composing for each node are comprised 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 guiding personnel to evacuate to the low node of weights from the high node of weights.So when the weights of comparison node, if ordinary node 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 first field also to compare, be to point to the neighbor node of layer or stair node upstairs or downstairs to determine this stair node.Therefore, the magnitude relationship of defined node weights: if to ordinary node w _x＜w _y, weight _x＞weight _yIf to stair node (l _x＞l _y) or (l _x=l _yAnd w _x＞w _y), weight _x＞weight _y

Initial phase divides two steps to complete: 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.Now node is described to the computation process of the jumping figure of Egress node as an example of the floor that comprises Egress node example, 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 respectively to export node i d, forward node id and forward node to the jumping figure of Egress node.At first one Egress node x broadcasts INIT message (x, x, 0), if node y judges that node x is that new Egress node is added into the outlet list after receiving this message, and records jumping figure h _xIf not new Egress node and h _x＞hop+1, h _x=hop+1, then reconstruct INIT message (x, y, h _x) and forward.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:

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

w _disRepresent the maximum weights that node can arrive, this value will be used to the weights of giving the node that disaster occurs.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, the jumping figure of nodal distance Egress node is less, and the initial weight of node will be less; 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 considers the reason of this parameter and be that the Egress node comprehensive distance index of node is identical, the less node of the jumping figure of nearest Egress node obviously larger node has better security.Parameter alpha is used for weighing initial weight and accounts for w _disProportion.If node type is outlet or stair node, its initial weight is zero.

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

After completing initial phase work, the initial weight of node and initial directional are as 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.The below is several key concepts of using in algorithm:

D a: constant.If the distance of certain sensor and any disaster origination point all is less than or equal to this constant, 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 fire detected, (4) weights of 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 that represents self.When Nodes was in the hazardous location, composing the floor hazard level to node was 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 simultaneously the urgent message of EMG (x, x, an A _dis, 0, type).This urgent message will be addressed in network, and other nodes are received and decided the channeling direction of self after urgent message by calculating.When a sensor node y receives urgent message EMG (x, z, the A that neighbor node z sends _z, hop, type) after, at first y uses A _zUpgrade the weights of z in neighbor list, then node y carries out according to the following steps:

1. node y judges whether x is new disaster information, comes new node more to the jumping figure e of disaster node 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 it is e _{X, y}Be updated to (hop+1); Otherwise do not make an amendment.

If z is the stair nodes, y be ordinary node y l is set _y=l _z, preventing 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 within representing that node y is positioned at the hazardous location, will raise immediately self weights of node y represent.If the lower floor section of ladder point of a stair node is in the hazardous location, think that these stair also are in the hazardous location, self weights also will raise.

(1) if y is ordinary node, and e _{X, y}≤ D arranges:

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 for gathering way of reflection weights, h _cWith h _dThe minor increment that represents respectively the nearest Egress node of present node and nearest disaster node.The weights that parameter μ is used for after restriction is upgraded 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.Obtain with max the maximum weights that can adjust in a plurality of disasters simultaneously.

(2) if y is the stair nodes, arrange by following rule:

If 1. $l_z^I<l_y^I$ And e _{X, y}-1≤D, 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 node y is in the hazardous location, arranges l _y=l _emg, w _yArrange according to formula (2);

3.y whether self become local minimum, that is to say that node y self is not 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 ordinary node, to its all neighbours x, w is arranged _y＜w _xIf y is the stair nodes, to its all neighbours x, A is arranged _y＜A _x

(1) if y is ordinary node, 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 allows continuously.

The implication of formula is as follows: when node becomes local minimum number of times continuously less than k, remove with the method for mean value the weights that raise; When node becomes local minimum number of times continuously more than or equal to k, show this Nodes in the situation that danger is surrounded, the weights that should raise more rapidly 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, 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, 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 the road that at this moment leads to other stair of floor is also covered by fire, just personnel should be guided seek other safe best-effort path to the upper strata.

If 1. $l_y=l_y^I,$ Y arranges 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 arranges l _y=l _dis, at this moment y judge it the l value of upper layer node whether less than l _dis, if y arranges w _y=0, at this moment because the lower floor section of ladder point of y is in the hazardous location, and also covered by the everybody towards the road with other stair of floor, so y can guide up floor ladder evacuation of personnel.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. if y meets the following conditions, continue broadcasting emergency message EMG (x, y, A _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 occurs _{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 ordinary node, node y points to the neighbours of weight w minimum in neighbours;

(2) if y is the stair nodes, node y points to the minimum neighbours of weights A (l, w) in neighbours.

Claims (3)

1. three-dimensional emergency escape guiding method based on wireless sensor network, it is characterized in that: adopt sensor network to monitor building building environment, when in building, the generation emergency need to be escaped, system can provide three-dimensional emergency escape guiding method, comprises the following 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 adjacent node is 1;

Initial phase divides two steps to complete:

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

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

(1.2) calculating of node initial weight

The account form of node initial weight is as follows:

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

w _disRepresent the maximum weights that node can arrive, this value will be used to the weights of giving the node that disaster occurs, 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 represents to weigh initial weight and accounts 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: if the distance of certain sensor and any disaster origination point all is less than or equal to this constant, claim this sensor to be positioned at the hazardous location;

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

The elemental height value of sensor 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 network simultaneously; After a sensor node y received the urgent message that neighbor node z sends, at first sensor node y used A _zUpgrade the height value of z in neighbor list, then node y carries out according to the following steps:

(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,yBe 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 it is e _x,yBe updated to (hop+1); Otherwise do not make an amendment;

(2.2) node y judgement e _x,yWhether be less than or equal to D, if set up within representing that node y is positioned at the hazardous location, raise immediately self weights of node y represent;

If (2.2.1) y is ordinary node, and e _x,y≤ D arranges 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 for gathering way of reflection weights, h _cWith h _dThe minor increment that represents respectively the nearest Egress node of present node and nearest disaster node, the weights that parameter μ is used for after restriction is upgraded exceed maximum weight w _dis,

The 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, arrange by following rule:

If 1.

And e _x,y-1≤D, 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. And e _x,y≤ D thinks that node y is in the hazardous location, arranges l _y=l _emg, w _yArrange 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 ordinary node, to its all neighbours x, w is arranged _y＜w _xIf y is the stair nodes, to its all neighbours x, A is arranged _y＜A _x

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

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 allows continuously;

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

If 1. Y arranges l _y=l _emg, w _yAdjust according to formula (3);

If 2. l _y=l _emg, y arranges l _y=l _dis, node y judges that whether its l value of upper layer node is less than l _dis, if so, node y arranges 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, continue the broadcasting emergency message:

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

If (2.4.2) e occurs to the jumping figure of disaster node x in node y _x,yChange;

(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 ordinary node, node y points to the neighbours of weight w minimum in neighbours;

If (2.5.2) y is the stair nodes, node y points to the minimum neighbours of weights A (l, w) in neighbours.

2. the three-dimensional 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 fire detected, 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 that represents self, when Nodes was in the hazardous location, composing the floor weights to node was l in algorithm _emg

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

Images