CN115719294A - Indoor pedestrian flow evacuation control method and system, electronic device and medium - Google Patents
Indoor pedestrian flow evacuation control method and system, electronic device and medium Download PDFInfo
- Publication number
- CN115719294A CN115719294A CN202211479239.2A CN202211479239A CN115719294A CN 115719294 A CN115719294 A CN 115719294A CN 202211479239 A CN202211479239 A CN 202211479239A CN 115719294 A CN115719294 A CN 115719294A
- Authority
- CN
- China
- Prior art keywords
- evacuation
- neural network
- density
- pedestrian flow
- pedestrian
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Alarm Systems (AREA)
Abstract
The invention discloses an indoor pedestrian flow evacuation control method, an indoor pedestrian flow evacuation control system, electronic equipment and an indoor pedestrian flow evacuation control medium, and relates to the field of pedestrian evacuation, wherein the method comprises the steps of firstly constructing a state equation and a measurement equation of outlet pedestrian density and a BP neural network for identifying a state function and an observation function and performing offline iterative training; solving a state function and an observation function according to the weight values of all layers in the BP neural network; predicting pedestrian flow evacuation density according to the solved function, and constructing an error online neural network for correcting a predicted value; and when the data state is abnormal, calculating the sum of the pedestrian flow evacuation density prediction value and the prediction error value as a pedestrian flow evacuation density prediction improvement value, and performing indoor pedestrian flow evacuation control according to the improvement value. The method can still ensure the accuracy of pedestrian flow evacuation density prediction under the condition of performance reduction or damage of the sensor, thereby improving the pedestrian flow evacuation efficiency.
Description
Technical Field
The present invention relates to the technical field of pedestrian evacuation, and in particular, to an indoor pedestrian flow evacuation control method, system, electronic device, and medium.
Background
With the continuous development of society, public places such as entertainment facilities and large shopping malls in various places of cities are increased rapidly. In the public places, people are often highly gathered, and treading events are easily caused after disasters such as fire disasters, earthquakes and the like occur. And in case of emergency, casualties can be greatly reduced if the guide personnel can be effectively evacuated. Crowd evacuation has therefore become a popular research topic.
Compared with the current research situation at home and abroad in China, the path search algorithm, the cellular automata model, the reinforcement learning control strategy and the like are widely applied to the research subject of pedestrian evacuation. However, a prerequisite for the above algorithm to perform well during crowd evacuation is the accurate measurement of sensors (such as cameras). If the influence of uncontrollable factors such as smoke and the like exists around the sensor at the moment, the distribution of the pedestrian may be measured inaccurately by the sensor, and therefore the working efficiency of the evacuation algorithm is influenced. In practical applications, the reduced efficiency of the evacuation algorithm may result in irreparable loss of lives and property.
Disclosure of Invention
The invention aims to provide an indoor pedestrian evacuation control method, an indoor pedestrian evacuation control system, an indoor pedestrian evacuation control electronic device and an indoor pedestrian evacuation control medium, so that the pedestrian evacuation efficiency under the condition that the performance of a sensor is reduced or damaged is improved.
In order to achieve the purpose, the invention provides the following scheme:
in one aspect, the invention provides an indoor pedestrian stream evacuation control method, which comprises the following steps:
modeling a pedestrian density evolution rule in an indoor evacuation scene, and constructing a state equation and a measurement equation of the pedestrian density at an outlet;
constructing a BP neural network for identifying a state function and an observation function in the state equation and the measurement equation; the BP neural network comprises an input layer, a hidden layer and an output layer;
performing offline iterative training on the BP neural network, and obtaining weights of all layers in the BP neural network after the training is finished;
solving a state function and an observation function in the state equation and the measurement equation according to the weight values of all layers in the BP neural network;
predicting pedestrian flow evacuation density according to the solved state function and the observation function to obtain a pedestrian flow evacuation density prediction value;
constructing an error online neural network for correcting the predicted pedestrian flow evacuation density value;
judging whether data state abnormity exists according to the state of the sensor in the indoor evacuation scene;
if the data state is not abnormal, indoor pedestrian flow evacuation control is directly carried out according to the pedestrian flow evacuation density predicted value, and the error online neural network is continuously trained;
if the data state is abnormal, calculating a prediction error value according to a currently trained error online neural network, and calculating the sum of the pedestrian flow evacuation density prediction value and the prediction error value to serve as a pedestrian flow evacuation density prediction improvement value;
and carrying out indoor pedestrian flow evacuation control according to the pedestrian flow evacuation density prediction improvement value.
Optionally, the modeling of the pedestrian density evolution rule in the indoor evacuation scene, and constructing a state equation and a measurement equation of the exit pedestrian density specifically include:
modeling a pedestrian density evolution rule in an indoor evacuation scene, and constructing a state equation X (k) = f (X (k-1)) + q (k-1) and a measurement equation Z (k) = h (X (k)) + r (k) of the exit pedestrian density; wherein the indoor evacuation scenario has a plurality of exits, and each exit is equipped with an evacuation guidance person and a sensor; x (k) is belonged to R n×n And X (k-1) represents the state change at time k and time k-1, respectivelyAn amount; z (k) is belonged to R m×n An observed variable representing time k; f (-) represents a nonlinear state function; h (-) represents a nonlinear observation function; q (k-1) is epsilon to R n×n Representing process noise; r (k) is belonged to R m×n Representing the measurement noise.
Optionally, the constructing a BP neural network for identifying a state function and an observation function in the state equation and the measurement equation specifically includes:
constructing a BP neural network for identifying a nonlinear state function f (·) in the state equation, wherein the BP neural network takes X (k) as an input, Z (k) as an output and a Sigmoid function as an activation function, and the function expression is Z (k) = h (X (k)) = W 2 T sigmoid(W 1 T X (k)); wherein W 1 As a weight of the input layer to the hidden layer, W 2 The weight from the hidden layer to the output layer;
constructing a BP neural network for identifying a nonlinear observation function h (·) in the measurement equation, wherein the BP neural network takes X (k-1) as an input, X (k) as an output and a Sigmoid function as an activation function, and the function expression is X (k) = f (X (k-1)) = W 2 T sigmoid(W 1 T X(k-1))。
Optionally, the performing offline iterative training on the BP neural network, and obtaining weights of each layer in the BP neural network after the training is finished, specifically includes:
obtaining sample data for training the BP neural network based on a pedestrian evacuation simulation experiment, and calculating the sample data by using a formula of 7: and 3, dividing the ratio into a training set and a verification set, updating the weight by taking SGD as an optimizer, taking MSE as a loss function, setting the learning rate and the training times to carry out iterative training, and obtaining the weight of each layer in the BP neural network after the training is finished.
Optionally, the constructing an error online neural network for correcting the predicted pedestrian flow evacuation density value specifically includes:
constructing an error online neural network for correcting the pedestrian flow evacuation density predicted value based on the BP neural network; the input of the error online neural network is a pedestrian flow evacuation density predicted value, and the output is an error between the pedestrian flow evacuation density predicted value and a pedestrian flow evacuation density true value, namely a prediction error value.
Optionally, the determining whether there is a data state abnormality according to the state of the sensor in the indoor evacuation scene specifically includes:
if the sensors in the indoor evacuation scene have performance degradation or damage states, determining that data state abnormity exists;
and if the sensors in the indoor evacuation scene do not have performance degradation or damage states, determining that no data state abnormity exists.
Optionally, the performing indoor pedestrian flow evacuation control according to the pedestrian flow evacuation density prediction improvement value specifically includes:
predicting and improving value rho of pedestrian stream evacuation density of ith exit at moment k i (k) Substituting into formula
Carrying out indoor pedestrian flow evacuation control; wherein
Representing the pedestrian flow evacuation target density of the ith exit at the k moment in the indoor evacuation scene; u. of i (k) The guiding action coefficient of the ith outlet at the k moment is represented; when u is i (k) Opening the guide of the ith outlet at 1, when u i (k) At 0, the guiding action of the ith outlet is closed.
On the other hand, the invention also provides an indoor pedestrian flow evacuation control system, which comprises:
the pedestrian density equation modeling module is used for modeling a pedestrian density evolution rule in an indoor evacuation scene and constructing a state equation and a measurement equation of the pedestrian density at an outlet;
the BP neural network construction module is used for constructing a BP neural network used for identifying state functions and observation functions in the state equation and the measurement equation; the BP neural network comprises an input layer, a hidden layer and an output layer;
the offline network iterative training module is used for performing offline iterative training on the BP neural network, and obtaining weights of all layers in the BP neural network after the training is finished;
the state and observation function solving module is used for solving state functions and observation functions in the state equation and the measurement equation according to the weights of all layers in the BP neural network;
the pedestrian flow evacuation density prediction module is used for predicting the pedestrian flow evacuation density according to the solved state function and the observation function to obtain a pedestrian flow evacuation density prediction value;
the error online neural network construction module is used for constructing an error online neural network for correcting the pedestrian flow evacuation density predicted value;
the data state abnormity judging module is used for judging whether data state abnormity exists according to the state of the sensor in the indoor evacuation scene;
the error online neural network training module is used for directly carrying out indoor pedestrian flow evacuation control according to the pedestrian flow evacuation density predicted value and continuously training the error online neural network if the data state abnormity does not exist;
the density prediction improvement value calculation module is used for calculating a prediction error value according to a currently trained error online neural network if the data state is abnormal, and calculating the sum of the pedestrian flow evacuation density prediction value and the prediction error value to be used as a pedestrian flow evacuation density prediction improvement value;
and the indoor pedestrian flow evacuation control module is used for carrying out indoor pedestrian flow evacuation control according to the pedestrian flow evacuation density prediction improvement value.
In another aspect, the present invention further provides an electronic device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the indoor pedestrian flow evacuation control method when executing the computer program.
In another aspect, the present invention also provides a non-transitory computer-readable storage medium having stored thereon a computer program which, when executed, implements the indoor pedestrian flow evacuation control method.
According to the specific embodiment provided by the invention, the invention discloses the following technical effects:
the invention provides an indoor pedestrian flow evacuation control method, an indoor pedestrian flow evacuation control system, electronic equipment and a medium, wherein the method comprises the following steps: modeling a pedestrian density evolution rule in an indoor evacuation scene, and constructing a state equation and a measurement equation of the pedestrian density at an outlet; constructing a BP neural network for identifying a state function and an observation function in the state equation and the measurement equation; the BP neural network comprises an input layer, a hidden layer and an output layer; performing offline iterative training on the BP neural network, and obtaining weights of all layers in the BP neural network after the training is finished; solving a state function and an observation function in the state equation and the measurement equation according to the weight values of all layers in the BP neural network; predicting pedestrian flow evacuation density according to the solved state function and the observation function to obtain a pedestrian flow evacuation density prediction value; constructing an error online neural network for correcting the pedestrian flow evacuation density predicted value; judging whether data state abnormity exists according to the state of the sensor in the indoor evacuation scene; if the data state is not abnormal, indoor pedestrian flow evacuation control is directly carried out according to the pedestrian flow evacuation density predicted value, and the error online neural network is continuously trained; if the data state is abnormal, calculating a prediction error value according to a currently trained error online neural network, and calculating the sum of the pedestrian flow evacuation density prediction value and the prediction error value to serve as a pedestrian flow evacuation density prediction improvement value; and carrying out indoor pedestrian flow evacuation control according to the pedestrian flow evacuation density prediction improvement value. The method can still ensure the accuracy of pedestrian flow evacuation density prediction under the condition of performance reduction or damage of the sensor, thereby improving the pedestrian flow evacuation efficiency.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings without creative efforts.
Fig. 1 is a flow chart of an indoor pedestrian evacuation control method according to the present invention;
fig. 2 is a technical route diagram of an indoor pedestrian stream evacuation control method according to the present invention;
fig. 3 is a schematic view of an indoor evacuation scenario provided by an embodiment of the present invention;
FIG. 4 is a schematic structural diagram of a BP neural network identified by a nonlinear observation function according to an embodiment of the present invention;
FIG. 5 is a schematic structural diagram of a BP neural network identified by a nonlinear state function according to an embodiment of the present invention;
fig. 6 is a schematic structural diagram of an error online neural network according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The invention aims to provide an indoor pedestrian flow evacuation control method, an indoor pedestrian flow evacuation control system, an indoor pedestrian flow evacuation control electronic device and an indoor pedestrian flow evacuation control medium, which can predict pedestrian flow evacuation density in a sensor (such as a camera) visual field on the basis of pedestrian flow evacuation density control simulation, so that the problem that the pedestrian flow evacuation density control cannot normally work after the performance of the sensor is reduced or even damaged (such as the camera is damaged due to fire) in the evacuation process is solved, and the pedestrian flow evacuation efficiency under the condition that the performance of the sensor is reduced or damaged is improved.
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.
Fig. 1 is a flow chart of an indoor pedestrian evacuation control method according to the present invention; fig. 2 is a technical route diagram of an indoor pedestrian flow evacuation control method according to the present invention. Referring to fig. 1 and 2, the present invention relates to an indoor pedestrian stream evacuation control method, including:
step 1: modeling is carried out on the pedestrian density evolution rule in the indoor evacuation scene, and a state equation and a measurement equation of the pedestrian density at the outlet are constructed.
Fig. 3 is a schematic view of an indoor evacuation scenario provided in an embodiment of the present invention. Referring to fig. 3, the evacuation scenario to which the present invention is directed is an indoor evacuation scenario, such as a mall, a casino, a library, a museum, etc., having a plurality of exits, each of which is equipped with evacuation guidance personnel and sensors; the sensor is typically a camera. The indoor evacuation scene is subjected to gridding processing, and a space comprising N multiplied by N grids can be formed. The embodiment shown in fig. 3 has 4 outlets, and each outlet is equipped with evacuation guidance personnel. Through the camera that is located the exit, evacuation guiding personnel can obtain the pedestrian density of the exit and adjust (open or close) the guide signal to realize the control to whole pedestrian flow evacuation guiding process.
The invention models the pedestrian density evolution rule in the indoor evacuation scene, and constructs the state equation and the measurement equation of the pedestrian density at the exit as follows.
The equation of state for exit pedestrian density is:
X(k)＝f(X(k-1))+q(k-1) (1)
the measurement equation for the exit pedestrian density is:
Z(k)＝h(X(k))+r(k) (2)
wherein X (k) is E.R n×n And X (k-1) ∈ R n×n Respectively representing state variables at the k moment and the k-1 moment; z (k) is belonged to R m×n An observed variable representing time k; f (-) represents a nonlinear state function; h (-) represents a nonlinear observation function; q (k-1) epsilon R n×n Representing process noise; r (k) is belonged to R m×n Representing the measurement noise.
Wherein r (k) and q (k-1) are independent of each other, have a mean value of 0 and satisfy the following equation:
E[r(k)]＝0,E[q(k-1)]＝0 (3)
E[r(k)r T (k)]＝R(k),E[q(k-1)q T (k-1)]＝Q(k-1) (4)
in the formula E [.]Indicating a desire; r (k) and Q (k-1) represent covariance matrices of R (k) and Q (k-1), respectively, and are both Gaussian white noise. Wherein Q (k-1) has a value of
The value of R (k) is 0.1.
In the embodiment shown in fig. 3, assuming that there is an error in the pedestrian density at the exit collected by the sensor due to the interference of unknown factors around the 4 th door (i.e., the exit 4), the pedestrian flow density detected by the sensor (camera) of the 4 th door corresponds to the observed variable Z (k), and the pedestrian flow densities detected by the sensors of the other three doors correspond to the state variable X (k):
z (k) = (density of pedestrian volume detected by the 4 th sensor) (6)
Because the state function f (-) and the observation function h (-) in the above equation can not be solved directly, the invention adopts BP neural network to identify, and then solve to obtain the nonlinear state function f (-) and the nonlinear observation function h (-).
And 2, step: and constructing a BP neural network for identifying the state function and the observation function in the state equation and the measurement equation.
For the embodiment shown in FIG. 3, in the equation of state (1) and measurement equation (2) for constructing the exit pedestrian density constructed in step 1, the state variable X (k-1) e R at time k-1 3×1 The state variable X (k) at the time k belongs to R 3×1 And the observed variable Z (k) at the moment k belongs to R 1×1 . Since the input and the output of the two functions f (-) and h (-) are different, two BP neural networks need to be constructed respectively for identifying the nonlinear state function f (-) and the nonlinear observation function h (-) respectively.Both BP neural networks comprise an input layer, an intermediate layer (typically a hidden layer) and an output layer.
The BP neural network structure for nonlinear observation function h (·) identification is shown in fig. 4, where dense _1_ input, dense _1, density and dense _2 represent the input, intermediate and output layers of the BP neural network, respectively; input and output represent the input and output of a certain layer of the BP neural network, respectively. In fig. 4, the people flow density of 1,2,3 gates at time k is used as input, so the input of the input layer is 3-dimensional, none represents the number of samples required to be input, and therefore the input of the input layer can be written as (None, 3); and (None, 3) is converted into (None, 7) by the input layer according to the connection weight of the input layer and the middle layer, and similarly, the middle layer obtains the output (None, 1) of the output layer through the connection weight, namely the people flow density of the 4 th door at the moment k. That is, the BP neural network for nonlinear observation function h (·) recognition of the present invention has X (k) as input, Z (k) as output, and Sigmoid function as activation function of the BP neural network. The functional expression of the BP neural network is therefore:
Z(k)＝h(X(k))＝W 2 T sigmoid(W 1 T X(k)) (7)
wherein W 1 As a weight of the input layer to the hidden layer, W 2 The weight from the hidden layer to the output layer; the Sigmoid function nonlinearizes the calculation result of linear weighting of each neuron in the neural network and gives the neural network the nonlinear mapping capability, and the expression of the function is
The BP neural network structure for nonlinear state function f (·) identification is shown in fig. 5, the BP neural network takes the human traffic density of 1,2,3 gates at the time k-1 as input, and the human traffic density of 1,2,3 gates at the time k as output, so the BP neural network structure shown in fig. 5 is different from fig. 4 in that in the output layer dense _2 of the BP neural network, the output dimension is 3 dimensions, i.e., the input layer converts (None, 3) into (None, 7) according to the connection weight of the input layer and the middle layer, and the middle layer obtains the output of the output layer as (None, 3) through the connection weight, i.e., the human traffic density of 1,2,3 gates at the time k. That is, the BP neural network for nonlinear state function f (-) recognition of the present invention has X (k-1) as input, X (k) as output, and Sigmoid function as the activation function of the BP neural network. The functional expression of the BP neural network is therefore:
X(k)＝f(X(k-1))＝W 2 T sigmoid(W 1 T X(k-1)) (8)
wherein W 1 As a weight of the input layer to the hidden layer, W 2 The weights from the hidden layer to the output layer.
And step 3: and performing offline iterative training on the BP neural network, and obtaining the weight of each layer in the BP neural network after the training is finished.
On the basis of the BP neural network constructed in the step 2, obtaining sample data for training the BP neural network based on a pedestrian evacuation simulation experiment, and calculating the sample data by a formula of 7:3, dividing the ratio into a training set and a verification set, updating the weight by taking SGD as an optimizer, taking MSE as a loss function, setting the learning rate and the training times to carry out iterative training, and obtaining the weight W of each layer in the BP neural network after the training is finished 1 And W 2 。
Specifically, a pedestrian evacuation simulation experiment based on the existing indoor multi-exit pedestrian flow simulation method is performed, during the operation of the pedestrian evacuation simulation experiment, pedestrian flow density data measured by four doors at each moment in the pedestrian evacuation process are recorded, and then 1000 times of simulation results are randomly selected and stored, so that sample data for training the BP neural network are obtained. Sample data was calculated as 7:3, dividing the ratio into a training set and a verification set, updating the weight by taking the SGD as an optimizer, taking the MSE as a loss function, and setting the learning rate and the training times for iterative training. And obtaining the weight of each layer after the BP neural network training is finished.
Wherein, the random gradient descent Strategy (SGD) adjusts the weight of the BP neural network in the negative gradient direction of the target gradient, and for a certain output node:
wherein eta represents the learning rate and takes the value of 0.01; e k Mean square error MSE; XW ij Represents the connection weight value between the ith layer and the jth layer of the BP neural network, delta W ij Connecting weight XW in training process ij The update amount of (2).
The connection weight and the threshold between the hidden layer and the input layer need to be adjusted as follows:
calculated connection weight XW ij Is a scalar; w 12 I.e. the connection weight between the input layer and the middle layer, and after the training is finished, the connection weight is used as W in the formulas (7) and (8) 1 ；W 23 That is, the connection weight between the middle layer and the output layer, and after training is finished, the connection weight is used as W in the formulas (7) and (8) 2 。W 1 、W 2 Are all in the form of a matrix.
In addition, in order to make the actual error and the prediction error with the network as small as possible, the weights of each layer of the BP neural network need to be continuously and iteratively adjusted according to a random gradient descent strategy, so that the value of the loss function is continuously reduced.
And 4, step 4: and solving a state function and an observation function in the state equation and the measurement equation according to the weight values of all layers in the BP neural network.
The weight W of each layer obtained by training 1 、W 2 Transfer function substituted into corresponding BP neural network
In the method, the state function and the observation function are respectively expressed in the formulas (7) and (8), so that the nonlinear state function f (-) and the nonlinear observation function h (-) can be obtained. In the formula W 1 As a connection weight of the input layer to the hidden layer, W 2 For the connection weight from the hidden layer to the output layer, input is the input of the BP neural network, and output is the output of the BP neural network.
After f (-) and h (-) are obtained, the people flow density of an outlet at each moment can be predicted in real time according to the formula (1) and the formula (2), and the pedestrian flow evacuation density prediction value of the corresponding outlet is obtained.
And 5: and predicting the pedestrian flow evacuation density according to the solved state function and the observation function to obtain a pedestrian flow evacuation density prediction value.
And predicting pedestrian flow evacuation density according to the solved nonlinear state function f (-) and the nonlinear observation function h (-) in the following steps:
1) Computing an a priori estimate X (k | k-1); the a priori estimate X (k | k-1) of the state variable X (k) at time k is calculated by substituting the state variable X (k-1) at time k-1 into equation (11):
X(k|k-1)＝f(X(k-1)) (11)
2) Predicting the pedestrian flow evacuation density predicted value Z of the 4 th exit at the k moment according to X (k | k-1) 4 (k):
Z 4 (k)＝H(k)X(k|k-1) (12)
Where H (k) is the Jacobian matrix of H (X (k | k-1)).
3) Computing covariance matrix P (k | k-1):
P(k|k-1)＝F(k|k-1)P(k-1)F T (k,k-1)+Q(k-1) (13)
wherein F (k | k-1) is a Jacobian matrix of F (X (k-1)); p (k-1) represents the covariance matrix at time k-1.
4) Calculating a gain K (K) for reflecting confidence degrees of the predicted value and the observed value:
K(k)＝P(k|k-1)H T (k)[H(k)P(k|k-1)H T (k)+R(k)] -1 (14)
5) Calculating a posterior estimate X (K) in combination with the gain K (K), i.e. correcting X (K | K-1):
X(k)＝X(k|k-1)+K(k)[Z(k)-h(X(k|k-1))] (15)
6) Calculating a covariance matrix P (k):
P(k)＝[I-K(k)H(k)]P(k|k-1) (16)
in the formula, I is a unit matrix; f (k | k-1) is a Jacobian matrix of F (X (k-1)); h (k) is H (X (k | k-1)))A jacobian matrix. The people flow density of a certain outlet at each moment can be recursively predicted according to the formulas (11) to (16), for example, the people flow evacuation density predicted value Z of the 4 th outlet at the moment k 4 (k) And a predicted pedestrian flow evacuation density value X (k) of 1,2,3 exits at the time k.
Step 6: and constructing an error online neural network for correcting the pedestrian flow evacuation density predicted value.
Namely, the input of the error online neural network constructed by the invention is the pedestrian flow evacuation density prediction value, in this embodiment, Z calculated by formula (12) 4 (k) (ii) a The output is the error between the predicted pedestrian flow evacuation density value and the true pedestrian flow evacuation density value, namely the predicted error value delta Z (k). Wherein the real value Z of the pedestrian flow evacuation density at the moment k True (k) Simulation experiment operation for pedestrian evacuationAnd recording the pedestrian flow density data of the 4 th door at the k moment in the pedestrian evacuation process. When there is no data state anomaly, Z 4 (k)＝Z True (k) (ii) a When there is a data state anomaly, Z True (k)＝Z 4 (k)+ΔZ(k)。
And 7: and judging whether the data state is abnormal or not according to the state of the sensor in the indoor evacuation scene.
The existing indoor multi-outlet pedestrian flow simulation method can work normally only if the pedestrian flow density information acquired by the sensor has higher accuracy. However, in an actual evacuation scene, for example, in an emergency evacuation scene from a fire, there may be a performance degradation condition of the sensor due to smoke shielding or a damage condition of the sensor due to the fire, and under the above conditions, the situation that the acquired data is inaccurate easily occurs in the camera for acquiring the evacuation density of people. The above problems cause that the existing method cannot continuously and normally work, thereby affecting the evacuation efficiency. In order to solve the problem, the invention proposes that the pedestrian density data acquired by the camera is firstly subjected to prediction correction to improve the accuracy of the data, and then indoor pedestrian flow evacuation control is carried out according to the corrected pedestrian flow evacuation density prediction improvement value, so that the influence of data errors on a guidance control system is reduced.
Specifically, if a sensor in the indoor evacuation scene has a performance degradation or damage state, determining that a data state abnormality exists; and if the sensors in the indoor evacuation scene do not have performance degradation or damage states, determining that no data state abnormity exists.
And 8: and if the data state is not abnormal, directly carrying out indoor pedestrian flow evacuation control according to the pedestrian flow evacuation density predicted value, and continuously training the error online neural network.
And if the data acquired by the current sensor is in a normal state, the error online neural network enters a training state, and the error online neural network is continuously trained. At this time, indoor pedestrian flow evacuation control can be directly carried out according to the pedestrian flow evacuation density predicted value, and a control algorithm is as follows:
but when judging that no data state abnormity exists, the predicted value Z of the pedestrian flow evacuation density of the ith exit at the moment k can be used i (k) (Z in the examples of the present invention) 4 (k) As ρ i (k) And substituting the formula (17) to carry out indoor pedestrian flow evacuation control. Where ρ is i aim Representing the pedestrian flow evacuation target density of the ith exit at the k moment in the indoor evacuation scene; u. u i (k) The guiding action coefficient of the ith outlet at the k moment is represented; when u is i (k) 1, opening the guide function of the ith outlet when u i (k) At 0, the guiding action of the ith outlet is closed. That is, the exit pedestrian density ρ at the time k obtained by the evacuation leading person at the ith exit i (k) Less than or equal to the set target density
While corresponding to the guiding function coefficient u of the door i (k) Will be adjusted to 1 and thus open the guiding action. Otherwise, the guiding function coefficient u of the corresponding door i (k) Will be adjusted to 0 and the guiding action will be switched off.
And step 9: and if the data state is abnormal, calculating a prediction error value according to the currently trained error online neural network, and calculating the sum of the pedestrian flow evacuation density prediction value and the prediction error value to serve as a pedestrian flow evacuation density prediction improvement value.
And if the current data acquired by the sensor is in an abnormal state, the error online neural network enters a use state at the moment, and a prediction error value delta Z (k) is calculated according to the currently trained error online neural network. And combining the network predicted value with the predicted error value obtained by the error online neural network to obtain an improved result. Namely, the pedestrian flow evacuation density prediction value Z is calculated according to the following formula (18) 4 (k) And the sum of the prediction error value delta Z (k) is used as a pedestrian flow evacuation density prediction improvement value Z Improvement of (k)：
Z Improvement of (k)＝Z 4 (k)+ΔZ(k) (18)
Step 10: and carrying out indoor pedestrian flow evacuation control according to the pedestrian flow evacuation density prediction improvement value.
When the data state is judged to be abnormal, the prediction improvement value Z of the pedestrian flow evacuation density of the ith exit at the moment k is used Improvement of (k) As ρ i (k) And substituting the formula (17) to carry out indoor pedestrian flow evacuation control. In the embodiment of the invention, when u 4 (k) 1, opening the 4 th outlet when u is 4 (k) At 0, the 4 th outlet was closed. The result shows that after the pedestrian density data acquired by the camera is corrected by the method, the indoor pedestrian flow evacuation control effect is remarkably improved.
The method comprises the steps of firstly modeling a pedestrian density evolution rule in an indoor evacuation scene, and constructing a nonlinear state function and a nonlinear observation function expression of the pedestrian density at an outlet; secondly, identifying and solving parameters of a nonlinear state function and a nonlinear observation function by adopting a BP neural network, and solving the problem that the function parameters cannot be directly solved through derivation; the predicted value is optimized and adjusted in real time through the error online neural network, and the prediction precision is further improved; and finally, the real-time prediction improvement value is used for pedestrian flow evacuation control, so that the evacuation efficiency under the condition of performance reduction or damage of the sensor is improved. The method of the invention realizes the prediction of pedestrian evacuation density in the visual field of the sensor (such as a camera) on the basis of the existing pedestrian flow evacuation density control simulation, thereby solving the problem that the pedestrian flow evacuation density control can not work normally after the performance of the sensor is reduced and even damaged (such as the damage of the camera caused by fire).
Based on the method provided by the invention, the invention also provides an indoor pedestrian flow evacuation control system, which comprises the following steps:
the pedestrian density equation modeling module is used for modeling a pedestrian density evolution rule in an indoor evacuation scene and constructing a state equation and a measurement equation of the pedestrian density at an outlet;
the BP neural network construction module is used for constructing a BP neural network used for identifying state functions and observation functions in the state equation and the measurement equation; the BP neural network comprises an input layer, a hidden layer and an output layer;
the offline network iterative training module is used for performing offline iterative training on the BP neural network, and obtaining weights of all layers in the BP neural network after the training is finished;
the state and observation function solving module is used for solving state functions and observation functions in the state equation and the measurement equation according to the weights of all layers in the BP neural network;
the pedestrian flow evacuation density prediction module is used for predicting the pedestrian flow evacuation density according to the solved state function and the observation function to obtain a pedestrian flow evacuation density prediction value;
the error online neural network construction module is used for constructing an error online neural network for correcting the pedestrian flow evacuation density predicted value;
the data state abnormity judging module is used for judging whether data state abnormity exists according to the state of the sensor in the indoor evacuation scene;
the error online neural network training module is used for directly carrying out indoor pedestrian flow evacuation control according to the pedestrian flow evacuation density predicted value and continuously training the error online neural network if the data state abnormity does not exist;
the density prediction improvement value calculation module is used for calculating a prediction error value according to a currently trained error online neural network if the data state is abnormal, and calculating the sum of the pedestrian flow evacuation density prediction value and the prediction error value to be used as a pedestrian flow evacuation density prediction improvement value;
and the indoor pedestrian flow evacuation control module is used for carrying out indoor pedestrian flow evacuation control according to the pedestrian flow evacuation density prediction improvement value.
On the whole, the method and the system adopt a modeling method to carry out prediction correction on pedestrian density data acquired by the camera, so that the accuracy of the data is improved; and then, indoor pedestrian stream evacuation control is performed according to the corrected data, so that the influence of data errors on a guide control system is reduced, and the evacuation efficiency of the sensor under the condition of performance reduction or damage is improved. The invention identifies the nonlinear state function and the nonlinear observation function training set by a neural network identification method, thereby solving the problems that the pedestrian flow evacuation system model is complex and is difficult to directly solve. The invention also adopts an off-line method to carry out the work of system identification, and because the identification data obtained in real time in the experiment is less, the off-line identification can fully utilize the historical data, thereby effectively improving the training efficiency of identification. The invention also adopts an error online neural network to correct the predicted data, if the model is only relied on to recur continuously, the error can be accumulated and amplified continuously along with the time, aiming at the problem, the invention utilizes the neural network to correct the predicted value data output by the model on line, thereby improving the robustness of the method in the practical application.
Further, the present invention also provides an electronic device, which may include: a processor, a communication interface, a memory, and a communication bus. The processor, the communication interface and the memory are communicated with each other through a communication bus. The processor may invoke a computer program in memory to perform the indoor pedestrian flow evacuation control method described.
Further, the computer program in the memory described above may be stored in a computer-readable storage medium when it is implemented in the form of a software functional unit and sold or used as a separate product. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: u disk, removable hard disk, read only memory, random access memory, magnetic or optical disk, etc. for storing program codes.
Further, the present invention also provides a non-transitory computer-readable storage medium having stored thereon a computer program which, when executed, may implement the indoor pedestrian flow evacuation control method.
In the present specification, the embodiments are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other. For the system disclosed by the embodiment, the description is relatively simple because the system corresponds to the method disclosed by the embodiment, and the relevant points can be referred to the method part for description.
The principles and embodiments of the present invention have been described herein using specific examples, which are provided only to help understand the method and the core concept of the present invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, the specific embodiments and the application range may be changed. In view of the above, the present disclosure should not be construed as limiting the invention.
Claims (10)
1. An indoor pedestrian flow evacuation control method is characterized by comprising the following steps:
modeling a pedestrian density evolution rule in an indoor evacuation scene, and constructing a state equation and a measurement equation of the pedestrian density at an outlet;
constructing a BP neural network for identifying a state function and an observation function in the state equation and the measurement equation; the BP neural network comprises an input layer, a hidden layer and an output layer;
performing offline iterative training on the BP neural network, and obtaining weights of all layers in the BP neural network after the training is finished;
solving a state function and an observation function in the state equation and the measurement equation according to the weight values of all layers in the BP neural network;
predicting pedestrian flow evacuation density according to the solved state function and the observation function to obtain a pedestrian flow evacuation density prediction value;
constructing an error online neural network for correcting the predicted pedestrian flow evacuation density value;
judging whether data state abnormity exists according to the state of the sensor in the indoor evacuation scene;
if the data state is not abnormal, indoor pedestrian flow evacuation control is directly carried out according to the pedestrian flow evacuation density predicted value, and the error online neural network is continuously trained;
if the data state is abnormal, calculating a prediction error value according to a currently trained error online neural network, and calculating the sum of the pedestrian flow evacuation density prediction value and the prediction error value to serve as a pedestrian flow evacuation density prediction improvement value;
and carrying out indoor pedestrian flow evacuation control according to the pedestrian flow evacuation density prediction improvement value.
2. The indoor pedestrian flow evacuation control method according to claim 1, wherein modeling pedestrian density evolution rules in an indoor evacuation scenario to construct an equation of state and a measurement equation of exit pedestrian density specifically comprises:
modeling a pedestrian density evolution rule in an indoor evacuation scene, and constructing a state equation X (k) = f (X (k-1)) + q (k-1) and a measurement equation Z (k) = h (X (k)) + r (k) of the exit pedestrian density; wherein the indoor evacuation scenario has a plurality of exits, and each exit is equipped with an evacuation guidance person and a sensor; x (k) is belonged to R n×n And X (k-1) represents state variables at time k and time k-1, respectively; z (k) is belonged to R m×n An observed variable representing time k; f (-) represents a nonlinear state function; h (-) represents a nonlinear observation function; q (k-1) is epsilon to R n×n Representing process noise; r (k) is belonged to R m×n Representing the measurement noise.
3. The indoor pedestrian flow evacuation control method according to claim 2, wherein the constructing a BP neural network for identifying a state function and an observation function in the state equation and the measurement equation specifically comprises:
constructed forIdentifying a BP neural network of a nonlinear state function f (·) in the state equation, wherein the BP neural network takes X (k) as input, Z (k) as output and a Sigmoid function as an activation function, and the function expression is Z (k) = h (X (k)) = W 2 T sigmoid(W 1 T X (k)); wherein W 1 As a weight of the input layer to the hidden layer, W 2 The weight from the hidden layer to the output layer;
constructing a BP neural network for identifying a nonlinear observation function h (·) in the measurement equation, wherein the BP neural network takes X (k-1) as an input, X (k) as an output and a Sigmoid function as an activation function, and the function expression is X (k) = f (X (k-1)) = W 2 T sigmoid(W 1 T X(k-1))。
4. The indoor pedestrian flow evacuation control method according to claim 3, wherein the performing offline iterative training on the BP neural network, and obtaining weights of each layer in the BP neural network after the training is finished, specifically comprises:
obtaining sample data for training the BP neural network based on a pedestrian evacuation simulation experiment, and calculating the sample data by using a formula of 7: and 3, dividing the ratio into a training set and a verification set, updating the weight by taking SGD as an optimizer, taking MSE as a loss function, setting the learning rate and the training times to carry out iterative training, and obtaining the weight of each layer in the BP neural network after the training is finished.
5. The indoor pedestrian flow evacuation control method according to claim 4, wherein the constructing an error online neural network for correcting the predicted pedestrian flow evacuation density value specifically comprises:
constructing an error online neural network for correcting the pedestrian flow evacuation density predicted value based on the BP neural network; the input of the error online neural network is a pedestrian flow evacuation density predicted value, and the output is an error between the pedestrian flow evacuation density predicted value and a pedestrian flow evacuation density true value, namely a prediction error value.
6. The indoor pedestrian flow evacuation control method according to claim 5, wherein the determining whether there is a data state anomaly according to the sensor state in the indoor evacuation scene specifically includes:
if the sensors in the indoor evacuation scene have performance degradation or damage states, determining that data state abnormity exists;
and if the sensors in the indoor evacuation scene do not have performance degradation or damage states, determining that no data state abnormity exists.
7. The indoor pedestrian stream evacuation control method according to claim 6, wherein the indoor pedestrian stream evacuation control according to the predicted pedestrian stream evacuation density improvement value specifically includes:
predicting and improving value rho of pedestrian stream evacuation density of ith exit at moment k i (k) Substitution formula
Carrying out indoor pedestrian flow evacuation control; wherein
Representing the pedestrian flow evacuation target density of the ith exit at the k moment in the indoor evacuation scene; u. of i (k) The guiding action coefficient of the ith outlet at the k moment is represented; when u is i (k) Opening the guide of the ith outlet at 1, when u i (k) At 0, the guiding action of the ith outlet is closed.
8. An indoor pedestrian flow evacuation control system, comprising:
the pedestrian density equation modeling module is used for modeling a pedestrian density evolution rule in an indoor evacuation scene and constructing a state equation and a measurement equation of the pedestrian density at an outlet;
the BP neural network construction module is used for constructing a BP neural network used for identifying state functions and observation functions in the state equation and the measurement equation; the BP neural network comprises an input layer, a hidden layer and an output layer;
the offline network iterative training module is used for performing offline iterative training on the BP neural network, and obtaining weights of all layers in the BP neural network after the training is finished;
the state and observation function solving module is used for solving state functions and observation functions in the state equation and the measurement equation according to the weights of all layers in the BP neural network;
the pedestrian flow evacuation density prediction module is used for predicting the pedestrian flow evacuation density according to the solved state function and the observation function to obtain a pedestrian flow evacuation density prediction value;
the error online neural network construction module is used for constructing an error online neural network for correcting the pedestrian flow evacuation density predicted value;
the data state abnormity judging module is used for judging whether data state abnormity exists according to the state of the sensor in the indoor evacuation scene;
the error online neural network training module is used for directly carrying out indoor pedestrian flow evacuation control according to the pedestrian flow evacuation density predicted value and continuously training the error online neural network if the data state abnormity does not exist;
the density prediction improvement value calculation module is used for calculating a prediction error value according to a currently trained error online neural network if the data state is abnormal, and calculating the sum of the pedestrian flow evacuation density prediction value and the prediction error value to be used as a pedestrian flow evacuation density prediction improvement value;
and the indoor pedestrian flow evacuation control module is used for carrying out indoor pedestrian flow evacuation control according to the pedestrian flow evacuation density prediction improvement value.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the indoor pedestrian flow evacuation control method according to any one of claims 1 to 7 when executing the computer program.
10. A non-transitory computer readable storage medium having stored thereon a computer program, wherein the computer program when executed implements the indoor pedestrian flow evacuation control method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211479239.2A CN115719294A (en) | 2022-11-23 | 2022-11-23 | Indoor pedestrian flow evacuation control method and system, electronic device and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211479239.2A CN115719294A (en) | 2022-11-23 | 2022-11-23 | Indoor pedestrian flow evacuation control method and system, electronic device and medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115719294A true CN115719294A (en) | 2023-02-28 |
Family
ID=85256150
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211479239.2A Pending CN115719294A (en) | 2022-11-23 | 2022-11-23 | Indoor pedestrian flow evacuation control method and system, electronic device and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115719294A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116385969A (en) * | 2023-04-07 | 2023-07-04 | 暨南大学 | Personnel gathering detection system based on multi-camera cooperation and human feedback |
| CN117311188A (en) * | 2023-09-26 | 2023-12-29 | 青岛理工大学 | Control method, system and equipment for crowd diversion railings in fixed places |
-
2022
- 2022-11-23 CN CN202211479239.2A patent/CN115719294A/en active Pending
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116385969A (en) * | 2023-04-07 | 2023-07-04 | 暨南大学 | Personnel gathering detection system based on multi-camera cooperation and human feedback |
| CN116385969B (en) * | 2023-04-07 | 2024-03-12 | 暨南大学 | Personnel gathering detection system based on multi-camera cooperation and human feedback |
| CN117311188A (en) * | 2023-09-26 | 2023-12-29 | 青岛理工大学 | Control method, system and equipment for crowd diversion railings in fixed places |
| CN117311188B (en) * | 2023-09-26 | 2024-03-12 | 青岛理工大学 | Control method, system and equipment for crowd diversion railings in fixed places |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107808122B (en) | Target tracking method and device | |
| CN115719294A (en) | Indoor pedestrian flow evacuation control method and system, electronic device and medium | |
| CN107145645B (en) | Method for predicting residual life of non-stationary degradation process with uncertain impact | |
| CN110245801A (en) | A kind of Methods of electric load forecasting and system based on combination mining model | |
| CN111612243A (en) | Traffic speed prediction method, system and storage medium | |
| CN110321603A (en) | A kind of depth calculation model for Fault Diagnosis of Aircraft Engine Gas Path | |
| JP2019061577A (en) | Abnormality determination method and program | |
| CN108304685A (en) | A kind of non-linear degradation equipment method for predicting residual useful life and system | |
| CN111047078B (en) | Traffic characteristic prediction method, system and storage medium | |
| CN112948743B (en) | Coal mine gas concentration deficiency value filling method based on space-time fusion | |
| CN110375736B (en) | Path planning method, system and device of intelligent device and readable storage medium | |
| CN111030889A (en) | Network traffic prediction method based on GRU model | |
| CN109598052B (en) | Intelligent ammeter life cycle prediction method and device based on correlation coefficient analysis | |
| CN110119540A (en) | A kind of multi output gradient promotion tree modeling method for survival risk analysis | |
| CN113988210A (en) | Method and device for restoring distorted data of structure monitoring sensor network and storage medium | |
| CN110795599B (en) | Video emergency monitoring method and system based on multi-scale graph | |
| CN111798494A (en) | Maneuvering target robust tracking method under generalized correlation entropy criterion | |
| CN112128950B (en) | Machine room temperature and humidity prediction method and system based on multiple model comparisons | |
| CN115938104A (en) | Dynamic short-time road network traffic state prediction model and prediction method | |
| CN112215495B (en) | Pollution source contribution calculation method based on long-time and short-time memory neural network | |
| CN109190749A (en) | A kind of prediction technique and device for the intelligent electric meter service life | |
| CN112098782A (en) | MOA insulation state detection method and system based on neural network | |
| CN116186581A (en) | Floor identification method and system based on graph pulse neural network | |
| CN112651168B (en) | Construction land area prediction method based on improved neural network algorithm | |
| CN114911157A (en) | Robot navigation control method and system based on partial observable reinforcement learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |