Road intelligent traffic multifunctional monitoring optimization system and method

Abstract

The invention discloses a road intelligent traffic multifunctional monitoring optimization system and method, and the system comprises a controller, a deceleration strip, and a traffic light assembly. The controller comprises an embedded single-chip microcomputer, a traffic flow detection module, a short-time traffic flow prediction module, a traffic state analysis module and a real-time traffic flow control adjustment module. The road intelligent traffic multifunctional monitoring optimization system not only has the capability of detecting and predicting the traffic flow in real time, but also has the function of monitoring the road traffic operation condition to give an alarm, and improves the travel experience of urban residents through intelligent regulation and control of the traffic flow. The technical problem that unnecessary waiting time or traffic jam occurs under some conditions due to the fact that traffic conditions cannot be intelligently regulated and controlled by an existing road intelligent traffic multifunctional monitoring optimization system and method, and traffic light time circulates in a fixed time interval is effectively solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent traffic lights, and in particular, relates to a multifunctional monitoring and optimization system and method for intelligent road traffic. Background Art

[0002] In the field of road traffic safety, a speed bump is a raised device installed on the road lane. It is a road safety traffic facility that forces vehicles to slow down. As we all know, speeding is one of the important causes of road traffic accidents, and the installation of speed bumps can reduce or avoid the occurrence of traffic accidents to a certain extent, playing a very important role in road traffic safety. With the continuous increase in urban traffic and the growth in the number of cars, traffic congestion has become one of the important problems facing today's society. Traditional traffic light systems are usually powered by the power grid. This power supply method has problems of energy waste and environmental pollution, and the traffic light time is cycled at fixed time intervals, which may cause unnecessary waiting time or traffic congestion in some cases. Therefore, there is a need for a multifunctional monitoring and optimization system and method for intelligent road traffic that monitors traffic information, predicts and reduces traffic congestion. Summary of the invention

[0003] The purpose of the present invention is to provide a road intelligent traffic multifunctional monitoring and optimization system and method for monitoring traffic information, predicting and reducing traffic congestion.

[0004] To achieve the above object, the present invention adopts the following technical solution:

[0005] A multifunctional monitoring and optimization system for intelligent road traffic, including a controller, speed bumps and traffic light components. The controller includes an embedded single-chip microcomputer, a vehicle flow detection module, a short-term vehicle flow prediction module, a traffic status analysis module and a real-time traffic flow control adjustment module;

[0006] The speed bump includes a left round head unit and a right round head unit which are arranged opposite to each other, and a plurality of interconnected middle units which are arranged between the left round head unit and the right round head unit;

[0007] The intermediate unit includes a base plate, a front block, a rear block, an upper cover, a hydraulic power generation mechanism, and a gear power generation mechanism. The front block and the rear block are respectively arranged at the front and rear ends of the base plate, and the hydraulic power generation mechanism and the gear power generation mechanism are both arranged on the base plate and located between the front block and the rear block.

[0008] The upper cover body includes a cover plate portion and a cylindrical portion, the upper surface of the cover plate portion is set as an arc surface, the lower surface of the cover plate portion is set as a plane, the upper end of the cylindrical portion is connected to the lower surface of the cover plate portion, the outer contour surface of the cylindrical portion is provided with a left-right symmetrical rack portion, a plurality of pressure power generation sheets are provided on the lower surface of the cylindrical portion, and a mounting hole is opened at the center position of the lower surface of the cylindrical portion; when a vehicle rolls over the upper cover body, pressure is applied to the pressure power generation sheet, so that the pressure power generation sheet generates electricity;

[0009] The base plate is provided with a mounting column corresponding to the mounting hole and a first return spring is provided on the mounting column, and the base plate and the upper cover body are assembled and installed by matching the mounting column and the mounting hole; there are two gear power generation mechanisms, which are respectively arranged on the left and right sides of the mounting column, and the gear power generation mechanism includes a primary gear shaft, a secondary gear shaft, a primary large gear, a secondary large gear, a primary small gear, a secondary small gear, a power generation gear, a gear generator, a front vertical plate, and a rear vertical plate; the front vertical plate and the rear vertical plate are arranged opposite to each other, and the front vertical plate and the rear vertical plate are provided with corresponding first upper holes, second upper holes, first lower holes, and second lower holes, and the primary gear shaft is installed on the first upper hole and the second upper hole through the first upper hole and the second upper hole The secondary gear shaft is installed at the upper part between the front vertical plate and the rear vertical plate through the first lower hole and the second lower hole, the first large gear is installed at the front end of the first large gear shaft, the first small gear is installed at the rear end of the first large gear shaft, the first small gear is meshed with the rack part, the second large gear is installed at the rear end of the second large gear shaft, the second small gear is installed at the front end of the second large gear shaft, the second small gear is meshed with the first large gear, the second large gear is meshed with the power generation gear, and the power generation gear is set on the output shaft of the gear generator through a key connection; the vehicle passes through the speed reduction belt, rolling over the upper cover body, so that the upper cover body moves downward, and then the rack part drives the gear generator to generate electricity through the gear transmission;

[0010] The electrical output terminal of the speed bump is connected to the traffic light assembly and the power storage box through electric wires, and the power storage box is connected to the traffic light assembly; the traffic light assembly includes a traffic light base, traffic light beads, and a moving wheel, the moving wheel is arranged at the bottom of the traffic light base, a traffic light frame is installed on the traffic light base, and the traffic light beads are installed on the traffic light frame.

[0011] The hydraulic power generation mechanism includes a liquid storage bag, hydraulic oil, a hydraulic generator and a liquid storage tank; the liquid storage bag is a ring-shaped liquid storage bag, which is filled with hydraulic oil. A circulation port is arranged on the side of the liquid storage bag, and the bag is connected to the hydraulic generator through a first circulation pipe. The hydraulic generator is connected to the liquid storage tank through a second circulation pipe. A plurality of second return springs are arranged in the liquid storage bag. When a vehicle passes through a speed bump, the liquid storage bag is squeezed, and the hydraulic oil enters the hydraulic generator through the first circulation pipe, converting the pressure of the hydraulic oil into electrical energy. The hydraulic oil enters the liquid storage tank through the second circulation pipe for short-term storage. When no vehicle passes by, the liquid storage bag is reset under the action of the second return spring, so that the hydraulic oil flows back from the liquid storage tank through the second circulation pipe and the first circulation pipe into the liquid storage bag for secondary power generation.

[0012] A first left pull ring and a first right pull ring are respectively arranged on the left and right sides of the lower surface of the upper cover body, a second left pull ring corresponding to the first left pull ring and a second right pull ring corresponding to the first right pull ring are arranged on the base plate, a first tension spring is installed between the first left pull ring and the second left pull ring, and a second tension spring is arranged between the first right pull ring and the second right pull ring.

[0013] A plurality of grooves arranged front to back are provided on the lower surfaces of the left round head unit, the middle unit and the right round head unit, and a plurality of heating wires are arranged in the bottom of the left round head unit, the middle unit and the right round head unit. When the heating wires are energized, the asphalt pavement is heated so that the uneven parts of the pavement are melted into the grooves to eliminate the gaps.

[0014] A first drainage ditch arranged on the left and right is provided at the connection between the base plate and the front block, a second drainage ditch arranged on the left and right is provided at the connection between the base plate and the rear block, a first drainage ditch connected to the first drainage ditch is provided on both the left and right sides of the front block, and a second drainage ditch connected to the second drainage ditch is provided on both the left and right sides of the rear block;

[0015] A long strip block is arranged on the mounting column, and the mounting holes are arranged with long strip holes at the front and rear, and a rotating hole is arranged in the cylindrical part, so that the long strip block enters the long strip hole, rotates and is supported, and the first return spring is supported by the lower surface of the cylindrical part, so that when the vehicle passes through the speed bump, the upper cover body compresses the first return spring downward, and the long strip block moves upward in the rotating hole;

[0016] The left round head unit and the right round head unit have the same structure, both comprising a connecting portion and a round head portion, wherein the connecting portion is arranged between the round head portion and the middle unit.

[0017] The road intelligent traffic multifunctional monitoring optimization method is implemented based on the above-mentioned road intelligent traffic multifunctional monitoring optimization system, and includes the following steps:

[0018] (1) Real-time monitoring of vehicle flow and type on the road;

[0019] (2) Predict short-term road vehicle flow based on LSTM;

[0020] (3) Analyze road traffic conditions;

[0021] (4) Make real-time intelligent adjustments to the duration of traffic lights to achieve intelligent adjustments for real-time traffic flow control.

[0022] The specific implementation process of step (1) is as follows:

[0023] ①Convert the pressure signal generated by the vehicle passing through the speed bump into an electrical signal;

[0024] ② Set different thresholds to distinguish different types of vehicles passing through speed bumps;

[0025] ③ Count and record the number of vehicles passing through the speed bump.

[0026] The specific implementation process of step (2) is as follows:

[0027] ① Preprocess the data by collecting historical traffic flow data of speed bumps on the road and the corresponding weather and holidays, setting the input variables as historical traffic flow, weather, and holidays, and the output variable as short-term traffic flow. The data is organized into the format of X = (X(1), X(2)...X(n)), where X represents the organized data set and n represents the number of samples;

[0028] The threshold method is used to process abnormal data. The data within the threshold is not deleted, and the deleted data is filled with the time series method. The threshold method is expressed as follows: 0≤X(j)≤θ, where X(j) represents the time series of traffic flow parameters at time j, which can refer to the traffic volume, weather, and holidays at time j; θ represents the threshold;

[0029] Traffic flow data is filtered according to weather records, where Vi represents the traffic volume under different weather conditions, Z represents the total traffic volume of the month, i represents different weather conditions, such as sunny, cloudy, light rain, moderate rain, heavy rain, snow, and fog, and C represents the traffic volume under different weather conditions. i Represents the influence coefficient of different weather conditions, θ i Represents the weather impact intensity factor, and sets the weather threshold to snowy and foggy daysθ i The value of .

[0030]

[0031] Traffic flow data is filtered according to weather records, where E r represents the traffic flow under different holidays, θ r Represents the holiday impact coefficient, r is the same as the holiday, such as Spring Festival, Qingming Festival, Labor Day, Dragon Boat Festival, Mid-Autumn Festival, National Day, set the holiday threshold to Spring Festival θ r Value

[0032]

[0033] If the data collected by the sensor at this moment exceeds the threshold, it is considered as erroneous data and is eliminated using the threshold method;

[0034] Use time series methods to handle missing data.

[0035]

[0036] Among them, α represents the number of data groups selected by the time series method, which is generally 2 or 3. The missing data can be supplemented by the time series method within a reasonable range;

[0037] ② Train the LSTM-based short-term traffic flow prediction model: divide the preprocessed traffic flow time series data set into two parts: a training set and a test set. The training set is used for modeling, and the test set is used to evaluate the model effect. Input the training set data into the model to train the model, and learn the characteristics of the data through the forget gate, input gate, and output gate of the LSTM, mainly learning the relationship between time.

[0038] f t =σ(W f β t +U f h t-1 +b f ) (4)

[0039] I t =σ(W I βt + U I h t-1 +b I )(5)

[0040] O t =σ(W o β t +U o h t-1 +b o )(6)

[0041] Initialize all weight matrices W, U and bias b, and perform forward propagation for each input training sample according to the formula to gradually calculate the forget gate f t , Input Gate I t , output gate O t The output values ​​of these gates are then used to calculate the new cell state S t , and substitute the calculated output value h t .

[0042] S t =ft S t-1 +I t (tanh(W S h t-1 +U S h t-1 )+b S )(7)

[0043] h t =O t tanhS t (8)

[0044] Among them, f t ,I t , O t are the output values ​​of the forget gate, input gate and output gate respectively, h t-1 is the hidden state vector of the LSTM unit at the previous moment; β t is the input vector of the LSTM unit at the current moment, is the historical traffic flow, weather and holiday characteristics; σ is the sigmoid activation function, W is the weight between the connected neurons, and U is the hidden state h at the previous moment. t-1 The weight matrix of the neuron, b is the bias of the neuron; S t is the new unit state vector obtained under the control of the forget gate and the update gate, combining the memory of the previous moment and the input data of this moment, t is the final output value under the control of the output gate;

[0045] The output value h is obtained by forward propagation of the LSTM model t , a label yi is needed to represent the real traffic flow data to calculate the loss. The loss function MAE formula is as follows:

[0046]

[0047] in, The average absolute error between the predicted and actual traffic volume is calculated, and then the weights and biases of the network are updated through back propagation. After a certain number of iterations, the loss converges and the training ends.

[0048] ③ Prediction performance evaluation and result analysis: Input the historical traffic flow, weather, holidays and other features in the test set into the loaded short-term traffic flow prediction model based on the LSTM neural network, update the model's weights to adapt to the data characteristics in the test set, save the weights, load the weight network, obtain the model's predicted output, and thus derive the predicted traffic flow. The predicted results are then input into the embedded MCU.

[0049] The specific implementation process of step (3) is as follows:

[0050] ① Install a laser speed meter at the edge of the road at one end of the speed bump, and the laser outlet of the laser rangefinder is flush with the ground. The principle is to generate a narrow laser beam through the internal laser transmitter, and emit it to the vehicle being tested that is passing through the speed bump, and measure the time it takes for the laser beam to be reflected by the vehicle surface and then return to the speed meter. This time is the flight time t of light. According to the constant speed of light and the time it takes for light to pass, the distance d between the vehicle being tested and the laser speed meter can be obtained.

[0051]

[0052] Among them, c represents the speed of light, which is 299792458 m / s in a vacuum; n represents the atmospheric refraction coefficient, which is about 1.0003;

[0053] During the speed measurement process, the laser speed meter will perform two laser ranging operations and record the time interval as the time difference T. By recording the distance moved by the measured vehicle during these two ranging processes, the average speed v of the vehicle during this period can be calculated.

[0054]

[0055] Where d1 and d2 represent the distance between the car and the speedometer during the first and second distance measurements respectively;

[0056] After obtaining the vehicle speed data, input it into the embedded microcontroller;

[0057] ② Determine the traffic status based on the vehicle speed data. The traffic status of the lane during the green light period:

[0058] When the vehicle passes through the speed bump, the speed v is close to or slightly lower than the speed limit specified by the road, and the lane is judged to be normal;

[0059] The speed v of the vehicle passing through the speed bump is far lower than the speed limit specified by the road, which is judged as lane congestion;

[0060] When the speed v of the vehicle passing through the speed bump is 0 for a long time and the stagnation time exceeds 15 minutes, it is judged as a lane accident;

[0061] Traffic operation status of the road in a traffic light cycle:

[0062] If all three lanes are normal, the road is considered normal;

[0063] If at least two of the three lanes are blocked, the road is considered blocked;

[0064] If one of the three lanes has an accident, it is considered a road accident;

[0065] ③ Alarm through the monitored traffic operation status. The embedded single chip computer transmits the monitored road traffic operation status to the corresponding traffic light through real-time data exchange and analysis to optimize the traffic light control and improve the road traffic efficiency. At the same time, it also transmits it to the wireless communication module. The wireless communication module uploads the road traffic operation status to the Internet for data sharing, realizes the sharing of road traffic information, and has an automatic alarm function. When the wireless communication module identifies a road accident, it will send the alarm information of the road accident to the transportation department in a timely and effective manner.

[0066] The specific implementation process of step (4) is as follows:

[0067] ① Based on the collected traffic flow data, the traffic light duration for each time period is pre-allocated. The traffic light is located in a two-way three-lane road. The four traffic phases are: Phase 1: East→West, West→East; Phase 2: East→South, West→North; Phase 3: North→South, South→North; Phase 4: South→West, North→East. The green light time for the first phase is S T1

[0068]

[0069] Among them, max(V x , V d ) represents the traffic flow in the 1-minute time window of the west-bound through lane and the traffic flow in the 1-minute time window of the east-bound through lane during the historical period of 7:30-9:30 from Monday to Friday, the maximum traffic flow, L is the distance from the speed reduction belt to the zebra crossing, v a is the maximum speed of the west-bound through lane and the east-bound through lane during this time period, and t0 is the buffer time, usually 1s;

[0070] The second phase green light time S T2

[0071]

[0072] Among them, max(V X , V D ) represents the traffic flow in the 1-minute time window of the west-bound left-turn lane and the traffic flow in the 1-minute time window of the east-bound left-turn lane during the historical period of 7:30-9:30 from Monday to Friday, the maximum traffic flow, L is the distance from the speed reduction belt to the zebra crossing, v b is the maximum speed of the west-bound left-turn lane and the east-bound left-turn lane during this time period, and t0 is the buffer time, usually 1s;

[0073] The third phase green light time S T3

[0074]

[0075] Among them, max(V b , V n ) represents the traffic flow in the 1-minute time window of the north-bound through lane and the traffic flow in the 1-minute time window of the south-bound through lane during the historical period of 7:30-9:30 from Monday to Friday, the maximum traffic flow, L is the distance from the speed reduction belt to the zebra crossing, v c is the maximum speed of the northbound through lane and the southbound through lane during this time period, and t0 is the buffer time, usually 1s;

[0076] Fourth phase green light time S T4

[0077]

[0078] Among them, max(V B , V N ) represents the traffic flow in the 1-minute window of the north-bound left-turn lane and the traffic flow in the 1-minute window of the south-bound left-turn lane during the historical period of 7:30-9:30 from Monday to Friday, the maximum traffic flow, L is the distance from the speed reduction belt to the zebra crossing, v c is the maximum speed of the northbound left-turn lane and the southbound left-turn lane during this time period, and t0 is the buffer time, usually 1s;

[0079] The first phase red light time S t1 =S T2 +S T3 +S T4 +9;

[0080] The second phase red light time S t2 =S T1 +S T3 +S T4 +9;

[0081] The third phase red light time S t3 =S T1 +S T2 +S T4 +9;

[0082] Fourth phase red light time S t4 =S T1 +S T2 +S T3 +9;

[0083] The total cycle of the traffic light in this time period is S T1 +S T2 +S T3 +S T4 +12, where each phase change is separated by a 3s yellow light;

[0084] ② Calculate the corresponding short-term traffic light duration based on the short-term traffic flow data predicted by LSTM, and compare it with the traffic light duration in the current time period. If there is a large gap, adjust the traffic light duration in the current time period according to the short-term traffic light duration. If there is only a small gap, do not make any adjustment and maintain the traffic light duration in the current time period.

[0085] Calculation of short-term traffic light time in the predicted future time period, first phase green light time ω T1

[0086]

[0087] in, represents the traffic flow in the 1-minute time window of the west-bound straight lane and the traffic flow in the 1-minute time window of the east-bound straight lane in the selected future time period, the maximum traffic flow, L is the distance from the speed reduction belt to the zebra crossing, v c is the maximum speed of the west-bound through lane and the east-bound through lane in the future time period, which is determined by The obtained,t0 is the buffer time, usually 1s;

[0088] Set the first phase green light time ω in the future time period T1 The first phase green light time S in the current time period T1 For comparison, set the difference threshold to 2s. T1 -S T1 |>2, then according to ω T1 Replace S with the value T1 The value of T1 -S T1 |≤2, then S is still maintained T1 By predicting the short-term traffic flow data in the future time period, the second phase green light time ω of the future time period is calculated. T2 , and the second phase green light time S in the current time period T2 Compare and calculate the second phase green light time ω in the future time period in turn T3 , compare the duration of the traffic lights in the four phases of the future time period with the duration of the traffic lights in the four phases of the current time period. If the difference exceeds the threshold, the duration of the traffic lights is adjusted. If the difference does not exceed the threshold, the duration of the traffic lights is not adjusted and the status quo is maintained;

[0089] The predicted short-term traffic flow data will be presented in chronological order and substituted into the calculation to continuously determine whether the traffic light duration needs to be adjusted. When the adjustment is made, the traffic light duration in the future time period will replace the traffic light duration in the current time period to complete the first regulation;

[0090] ③ Through real-time monitoring of the three traffic operation states of normal lane, lane congestion, and lane accident, set a second control strategy for the duration of traffic lights in different states;

[0091] When the lane is normal, keep the traffic light duration after the first adjustment to ensure smooth traffic flow;

[0092] In the case of lane congestion, determine the phase corresponding to the lane, extend the green light time of the phase after the first adjustment by 2s, and shorten the red light time by 2s, that is, shorten the green light time of other phases without lane congestion after the first adjustment by an average of 2s in total. If lanes in all four phases are congested, only the green light time will be extended and the red light time will not be shortened. When multiple lanes are congested, the lanes to which the phase belongs have If a lane is blocked, the green light time after the first adjustment of this phase will be extended Shortened red light time

[0093] In the event of a lane accident, the phase corresponding to the lane is determined, and the green light time after the first adjustment of that phase is set to 0, and the red light time is extended indefinitely and keeps flashing in an alarm state until the alarm is lifted. The traffic light duration after the first adjustment of that phase can be restored to normal to avoid further traffic congestion and accident risks.

[0094] The speed bump of the road intelligent traffic multifunctional monitoring and optimization system of the present invention is provided with a gear power generation mechanism. When a vehicle passes through the speed bump, it rolls over the upper cover body, causing the upper cover body to move downward, thereby causing the rack part to drive the gear generator to generate electricity through gear transmission, and a hydraulic power generation mechanism is provided. When a vehicle passes through the speed bump, the liquid storage bag is squeezed, and the hydraulic oil enters the hydraulic generator through the first circulation pipeline, converting the pressure of the hydraulic oil into electrical energy, and the hydraulic oil enters the liquid storage tank through the second circulation pipeline for short-term storage. When no vehicle passes, the liquid storage bag is reset under the action of the second reset spring, so that the hydraulic oil flows from the liquid storage tank through the second circulation pipeline and the first circulation pipeline back to the liquid storage bag for secondary power generation. While reducing the speed of the vehicle, it also realizes the reuse and conservation of energy, which can reduce the dependence on traditional energy and reduce the energy cost of the city.

[0095] The road intelligent traffic multifunctional monitoring and optimization system of the present invention not only has the ability to detect and predict traffic flow in real time, but also has the function of monitoring road traffic operation conditions and issuing alarms, and improves the travel experience of urban residents by intelligently regulating traffic flow.

[0096] In summary, the present invention can intelligently control traffic conditions, improve the travel experience of urban residents, and effectively solve the technical problem that the existing road intelligent traffic multifunctional monitoring and optimization system and method cannot intelligently control traffic conditions, resulting in the traffic light time being cycled within a fixed time interval, further leading to unnecessary waiting time or traffic congestion in some cases. BRIEF DESCRIPTION OF THE DRAWINGS

[0097] Figure 1 It is a structural schematic diagram of the traffic light device of the present invention.

[0098] Figure 2 It is a structural schematic diagram of the speed reduction belt of the present invention.

[0099] Figure 3 It is a structural schematic diagram of the intermediate unit of the speed bump of the present invention.

[0100] Figure 4 It is a schematic diagram of the internal structure of the intermediate unit of the present invention.

[0101] Figure 5 It is a structural schematic diagram of the gear power generation mechanism of the present invention.

[0102] Figure 6 It is a structural schematic diagram of the upper cover body of the present invention.

[0103] Figure 7 It is a structural schematic diagram of the left round head unit of the present invention.

[0104] Figure 8 It is a schematic diagram of the structure of the bottom groove of the intermediate unit of the present invention.

[0105] Fig. 9 It is a schematic diagram of the module composition of the controller of the present invention.

[0106] Fig.10 It is a flow chart of the method for real-time monitoring of vehicle flow of the present invention.

[0107] Fig.11 It is a flow chart of the present invention for predicting short-term road traffic flow based on LSTM.

[0108] Fig.12 It is a flow chart of the method for monitoring the traffic operation status on the road of the present invention.

[0109] Fig.13 It is a flow chart of the method for real-time intelligent control of traffic light duration of the present invention. DETAILED DESCRIPTION

[0110] The embodiments of the present invention are further described below with reference to the accompanying drawings.

[0111] Embodiment 1

[0112] like Figure 1-13 As shown, a multifunctional monitoring and optimization system for intelligent road traffic includes a controller, a speed bump 40 and a traffic light assembly 41. The controller includes an embedded single-chip microcomputer, a vehicle flow detection module, a short-term vehicle flow prediction module, a traffic state analysis module and a real-time traffic flow control adjustment module.

[0113] It includes a left round head unit 1 and a right round head unit 2 which are arranged opposite to each other, and a plurality of interconnected middle units 3 which are arranged between the left round head unit 1 and the right round head unit 2;

[0114] The intermediate unit 3 includes a base plate 4, a front block 5, a rear block 6, an upper cover 7, a hydraulic generator 26 structure, and a gear generator 22 structure. The front block 5 and the rear block 6 are respectively arranged at the front and rear ends of the base plate 4, and the hydraulic generator 26 structure and the gear generator 22 structure are both arranged on the base plate 4 and located between the front block 5 and the rear block 6.

[0115] The upper cover body 7 includes a cover portion 8 and a cylindrical portion 9. The upper surface of the cover portion 8 is set as an arc surface, and the lower surface of the cover portion 8 is set as a plane. The upper end of the cylindrical portion 9 is connected to the lower surface of the cover portion 8. A left-right symmetrical rack portion 10 is set on the outer contour surface of the cylindrical portion 9. A plurality of pressure generating sheets 11 are set on the lower surface of the cylindrical portion 9. A plurality of grooves are set on the lower surface of the cylindrical portion 9. The pressure generating sheets 11 are set in the grooves, and there are 40 pressure generating sheets 11 in total. Each pressure generating sheet 11 corresponds to a groove. After being subjected to pressure from the upper cover body 7, these pressure generating sheets 11 will produce deformation and then stimulate the piezoelectric effect inside them, converting mechanical energy into electrical energy, which is transmitted to the external circuit through its electrical output terminal. Part of it is used for power supply and part of it is stored, thereby realizing energy recovery and reuse. A mounting hole 12 is provided at the center of the lower surface of the cylindrical portion 9; when the vehicle rolls over the upper cover 7, pressure is applied to the pressure generating sheet 11, so that the pressure generating sheet 11 generates electricity; the number of the pressure generating sheets 11 can be adjusted as needed to adapt to different pressures and power requirements, and the pressure generating sheets 11 are installed inside the speed bump 40, protected, and can adapt to various environmental conditions to ensure the stable operation of the system; a mounting column 13 corresponding to the mounting hole 12 is provided on the base plate 4, and a first reset spring is provided on the mounting column 13 Spring 14, through the cooperation of mounting column 13 and mounting hole 12, the base plate 4 and the upper cover body 7 are assembled and installed; there are two gear generators 22, which are respectively arranged on the left and right sides of the mounting column 13, and the gear generator 22 includes a primary gear shaft 15, a secondary gear shaft 16, a primary large gear 17, a secondary large gear 18, a primary small gear 19, a secondary small gear 20, a power generation gear 21, a gear generator 22, a front vertical plate 23, and a rear vertical plate 24; the front vertical plate 23 and the rear vertical plate 24 are arranged oppositely, and the front vertical plate 23 and The rear vertical plate 24 is provided with corresponding first upper holes, second upper holes, first lower holes, and second lower holes. The first-stage gear shaft 15 is installed at the upper part between the front vertical plate 23 and the rear vertical plate 24 through the first upper holes and the second upper holes. The second-stage gear shaft 16 is installed at the lower part of the front vertical plate 23 and the rear vertical plate 24 through the first lower holes and the second lower holes. The first-stage large gear 17 is installed at the front end of the first-stage gear shaft 15, and the first-stage small gear 19 is installed at the rear end of the first-stage gear shaft 15. The first-stage small gear 19 is meshed with the rack portion 10. The secondary large gear 18 is installed at the rear end of the secondary gear shaft 16, and the secondary small gear 20 is installed at the front end of the secondary gear shaft 16. The secondary small gear 20 meshes with the primary large gear 17 to achieve power transmission and amplification of the transmission ratio. The secondary large gear 18 meshes with the power generation gear 21, and the power generation gear 21 is set on the output shaft of the gear generator 22 through a key connection; the vehicle passes through the speed reduction belt 40, rolling over the upper cover 7, so that the upper cover 7 moves downward, and then the rack part 10 drives the gear generator 22 to generate electricity through the gear transmission. As the rotor rotates, the magnetic field in the gear generator 22 changes, cutting the magnetic flux lines, thereby generating current in the coil of the generator.The gear ratio is amplified by the gear configuration on the primary gear shaft 15 and the secondary gear shaft 16, and the rotation speed is increased to meet the input requirements of the generator. The mechanism of the gear generator 22 converts the mechanical energy of the vehicle passing through the speed bump 40 into electrical energy, and its electrical output terminal is connected to an external circuit to transmit the generated electrical energy to where it is needed, such as a traffic light or an energy storage system.

[0116] The hydraulic generator 26 comprises a liquid storage bag 25, hydraulic oil, a hydraulic generator 26 and a liquid storage tank 27; the liquid storage bag 25 is an annular liquid storage bag 25, which is filled with hydraulic oil. A flow port is arranged on the side of the liquid storage bag 25, and the liquid storage bag 25 is connected to the hydraulic generator 26 through a first flow pipe, and the hydraulic generator 26 is connected to the liquid storage tank 27 through a second flow pipe. A plurality of second return springs are arranged in the liquid storage bag 25. When a vehicle passes through the speed bump 40, the liquid storage bag 25 is squeezed, and the hydraulic oil enters the hydraulic generator 26 through the first flow pipe, and the pressure of the hydraulic oil is converted into electrical energy. The hydraulic oil enters the liquid storage tank 27 through the second flow pipe for short-term storage. When no vehicle passes, the liquid storage bag 25 is reset under the action of the second return spring, so that the hydraulic oil flows from the liquid storage tank 27 through the second flow pipe and the first flow pipe back to the liquid storage bag 25 for secondary power generation. After the vehicle passes, the spring inside the liquid storage bag 25 helps it to reset quickly, ready to respond to the next pressure change. The design of the liquid storage bag 25 ensures a quick response to the pressure changes when the vehicle passes through the speed bump 40, and quickly resets after the pressure disappears. The setting of the liquid storage tank 27 allows the hydraulic oil to be stored briefly after flowing through the hydraulic generator 26, and to achieve secondary power generation when it flows back to the liquid storage bag 25, increasing the efficiency of energy recovery. The closed-loop design of the hydraulic system not only improves energy utilization, but also facilitates system maintenance and monitoring. The electrical energy generated by the hydraulic generator 26 can be transmitted to the external circuit through the electrical output terminal for use by traffic lights, street lights or energy storage systems. The entire hydraulic generator 26 structure is installed inside the block base of the speed bump 40, reducing the impact of external environmental factors and protecting the stability and durability of the system.

[0117] A first left pull ring 28 and a first right pull ring 29 are respectively provided on the left and right sides of the lower surface of the upper cover body 7, a second left pull ring 30 corresponding to the first left pull ring 28 and a second right pull ring corresponding to the first right pull ring 29 are provided on the base plate 4, a first tension spring 31 is installed between the first left pull ring 28 and the second left pull ring 30, and a second tension spring is provided between the first right pull ring 29 and the second right pull ring.

[0118] A plurality of grooves 32 arranged front to back are provided on the lower surfaces of the left round head unit 1, the middle unit 3 and the right round head unit 2. A plurality of heating wires are arranged in the bottom of the left round head unit 1, the middle unit 3 and the right round head unit 2. When the heating wires are energized, the asphalt pavement is heated so that the uneven parts of the pavement are melted into the grooves 32 to eliminate the gaps.

[0119] The connection between the base plate 4 and the front block 5 is provided with a first drainage ditch 33 arranged on the left and right, and the connection between the base plate 4 and the rear block 6 is provided with a second drainage ditch 34 arranged on the left and right. The left and right sides of the front block 5 are provided with a first drainage groove 35 connected to the first drainage ditch 33, and the left and right sides of the rear block 6 are provided with a second drainage groove 36 connected to the second drainage ditch 34. This design is intended to effectively manage and guide the water flow entering from the gap of the upper cover 7 of the speed bump 40, ensure that the water flow can be discharged smoothly, avoid the accumulation of water to damage the speed bump 40 or affect its performance, and can efficiently collect and guide the water flow. The first drainage ditch 33, the first drainage groove 35, the second drainage groove 34 and the second drainage groove 36 allow the water flow to flow in smoothly, further guide the water flow through the inside of the base, form a continuous drainage path, and effectively ensure that the water flow can flow smoothly from the gap of the upper cover 7 of the speed bump 40 into the first drainage ditch 33 and the second drainage ditch 34, and then be discharged through the first drainage groove 35 and the second drainage groove 36, thereby achieving an effective drainage effect.

[0120] A long strip block 37 is provided on the mounting column 13, and the mounting hole 12 is provided with long strip holes at the front and rear. A rotating hole is provided in the cylindrical portion 9, so that the long strip block 37 enters the long strip hole and rotates and is supported. The first return spring 14 is supported by the lower surface of the cylindrical portion 9, so that when the vehicle passes through the speed bump 40, the upper cover body 7 compresses the first return spring 14 downward, and the long strip block 37 moves upward in the rotating hole.

[0121] The left round head unit 1 and the right round head unit 2 have the same structure, both comprising a connecting portion 38 and a round head portion 39, wherein the connecting portion 38 is arranged between the round head portion 39 and the middle unit 3. The traffic light assembly 41 comprises a traffic light base, a traffic light lamp bead, and a moving wheel, wherein the moving wheel is arranged at the bottom of the traffic light base, a traffic light lamp holder is installed on the traffic light base, and the traffic light lamp bead is installed on the traffic light lamp holder.

[0122] Embodiment 2

[0123] The road intelligent traffic multifunctional monitoring optimization method is implemented based on the above-mentioned road intelligent traffic multifunctional monitoring optimization system, and includes the following steps:

[0124] (1) Real-time monitoring of vehicle flow and type on the road;

[0125] (2) Predict short-term road vehicle flow based on LSTM;

[0126] (3) Analyze road traffic conditions;

[0127] (4) Make real-time intelligent adjustments to the duration of traffic lights to achieve intelligent adjustments for real-time traffic flow control.

[0128] The specific implementation process of step (1) is as follows:

[0129] ① Convert the pressure signal generated by the vehicle passing through the speed bump 40 into an electrical signal; the speed bump 40 is installed at the road intersection. When the tire of the vehicle passes through the speed bump 40 at the road intersection, the weight of the vehicle will be applied to the surface of the speed bump 40, causing the speed bump 40 to be under pressure. The pressure sensor inside the speed bump 40 is set on the surface of the pressure generating sheet. When the pressure sensor senses this pressure, it converts it into an electrical signal through the pressure sensitive element. Subsequently, this electrical signal is captured by the circuit inside the pressure sensor and amplified, processed and recorded. In the actual working process, considering the complex traffic conditions at the intersection and the need for a stable and reliable monitoring system, this embodiment selects a piezoresistive sensor. This sensor has high durability and stability, can work for a long time under variable environmental conditions without being affected, and is easier to install and maintain. In this embodiment, considering the situation where multiple vehicles pass through the speed bump 40 at the same time, in order to ensure the accuracy of vehicle flow monitoring, a strategy of installing multiple piezoresistive sensors on the speed bump 40 is adopted. Each sensor is positioned at a specific position on the speed bump 40 to ensure that it is associated with a specific lane. Each sensor can independently monitor and record vehicles passing through its lane.

[0130] In this embodiment, considering the tire layout of the vehicle, including the front tire, the rear tire and the middle tire of the three-axle vehicle, an electrical signal will be generated when passing through the speed bump 40. By ensuring the synchronization of the front and rear tire signals, it is judged whether a vehicle has passed through the speed bump 40. When a vehicle passes through the speed bump 40, the time interval between the front and rear tires generating electrical signals is relatively small, about 0.1 seconds to 1.5 seconds. If the time interval between the detected signals is too large and is not within this range, it is a different vehicle. When the vehicle stops at a red light, the counting of the vehicle passing through the speed bump 40 can be suspended. When the green light comes, continue to monitor the situation of the vehicle passing through the speed bump 40. This can avoid misjudgment caused by stopping and waiting for the red light.

[0131] ② Different thresholds are set to distinguish different types of vehicles passing through the speed bump 40; the acquired electrical signals are amplified, filtered and analyzed to extract the signal characteristics that best reflect the passage of different types of vehicles, that is, the pressure change amplitude A. Different types of vehicles passing through the speed bump 40 may produce pressure changes of different amplitudes. Larger vehicles (such as trucks) may produce larger pressure changes, while smaller vehicles (such as cars) may produce smaller pressure changes.

[0132] Using a support vector machine (SVM) algorithm, models of different types of vehicles are trained based on the pressure variation amplitude A extracted from the electrical signals converted when different types of vehicles pass through the speed bump 40. According to the trained model, thresholds of different pressure variation amplitudes A are set to distinguish vehicle types.

[0133] The preset pressure variation amplitude A is within the range of 0.1 Pa to 0.4 Pa, and the passing vehicle is determined to be a small vehicle (sedan, small truck).

[0134] The preset pressure variation amplitude A is within the range of 0.4 Pa to 0.7 Pa, and the passing vehicle is determined to be a medium-sized vehicle (medium-sized truck, SUV).

[0135] The preset pressure change amplitude A is within the range of 0.7 Pa to 1 Pa, and the passing vehicle is determined to be a large vehicle (large truck, bus).

[0136] ③ Count and record the vehicles passing through the speed bump 40. By monitoring the electrical signal converted by the piezoresistive sensor, it is determined whether a vehicle has passed. When the time interval between two consecutive electrical signals is less than 1.5 seconds, it is counted as a vehicle passing. A specific threshold of the pressure change amplitude A is set to classify the vehicles passing through the speed bump 40. According to the time series, the time is divided into 1-minute time windows. Using the cumulative counting method, the number of vehicles passing through the speed bump 40 in each time window will be classified and accumulated. In this way, the vehicles passing through the speed bump 40 are classified and counted, and the traffic flow on the road is detected. The traffic flow data is input into the embedded single-chip microcomputer for recording.

[0137] The specific implementation process of step (2) is as follows:

[0138] ① Preprocess the data by collecting the historical traffic flow data of 40 speed bumps on the road and the corresponding weather and holidays, setting the input variables as historical traffic flow, weather, and holidays, and the output variable as short-term traffic flow. The data is organized into the format of X = (X(1), X(2)...X(n)), where X represents the organized data set and n represents the number of samples;

[0139] The threshold method is used to process abnormal data. The data within the threshold is not deleted, and the deleted data is filled with the time series method. The threshold method is expressed as follows: 0≤X(j)≤θ, where X(j) represents the time series of traffic flow parameters at time j, which can refer to the traffic volume, weather, and holidays at time j; θ represents the threshold;

[0140] Traffic flow data is filtered according to weather records, where Vi represents the traffic volume under different weather conditions, Z represents the total traffic volume of the month, i represents different weather conditions, such as sunny, cloudy, light rain, moderate rain, heavy rain, snow, and fog, and C represents the traffic volume under different weather conditions. i Represents the influence coefficient of different weather conditions, θ i Represents the weather impact intensity factor, and sets the weather threshold to snowy and foggy daysθ i The value of .

[0141]

[0142] Traffic flow data is filtered according to weather records, where E r represents the traffic flow under different holidays, θ r Represents the holiday impact coefficient, r is the same as the holiday, such as Spring Festival, Qingming Festival, Labor Day, Dragon Boat Festival, Mid-Autumn Festival, National Day, set the holiday threshold to Spring Festival θ r Value

[0143]

[0144] If the data collected by the sensor at this moment exceeds the threshold, it is considered as erroneous data and is eliminated using the threshold method;

[0145] Use time series methods to handle missing data.

[0146]

[0147] Among them, α represents the number of data groups selected by the time series method, which is generally 2 or 3. The missing data can be supplemented by the time series method within a reasonable range;

[0148] ② Train the LSTM-based short-term traffic flow prediction model: divide the preprocessed traffic flow time series data set into two parts: a training set and a test set. The training set is used for modeling, and the test set is used to evaluate the model effect. Input the training set data into the model to train the model, and learn the characteristics of the data through the forget gate, input gate, and output gate of the LSTM, mainly learning the relationship between time.

[0149] f t =σ(W f β t +U f h t-1 +b f ) (4)

[0150] I t =σ(W I β t +U I h t-1 +b I )(5)

[0151] O t =σ(W o β t +U o h t-1 +b o )(6)

[0152] Initialize all weight matrices W, U and bias b, and perform forward propagation for each input training sample according to the formula to gradually calculate the forget gate f t , Input Gate I t , output gate O t The output values ​​of these gates are then used to calculate the new cell state S t , and substitute the calculated output value h t .

[0153] S t =f t S t-1 +I t (tanh(W S h t-1 +U S h t-1 )+b S )(7)

[0154] h t =O t tanhS t (8)

[0155] Among them, f t ,I t , O t are the output values ​​of the forget gate, input gate and output gate respectively, h t-1 is the hidden state vector of the LSTM unit at the previous moment; β t is the input vector of the LSTM unit at the current moment, is the historical traffic flow, weather and holiday characteristics; σ is the sigmoid activation function, W is the weight between the connected neurons, and U is the hidden state h at the previous moment. t-1 The weight matrix of the neuron, b is the bias of the neuron; S t is the new unit state vector obtained under the control of the forget gate and the update gate, combining the memory of the previous moment and the input data of this moment, t is the final output value under the control of the output gate;

[0156] The output value h is obtained by forward propagation of the LSTM model t , a label yi is needed to represent the real traffic flow data to calculate the loss. The loss function MAE formula is as follows:

[0157]

[0158] in, The average absolute error between the predicted and actual traffic volume is calculated, and then the weights and biases of the network are updated through back propagation. After a certain number of iterations, the loss converges and the training ends.

[0159] ③ Prediction performance evaluation and result analysis: Input the historical traffic flow, weather, holidays and other features in the test set into the loaded short-term traffic flow prediction model based on the LSTM neural network, update the model's weights to adapt to the data characteristics in the test set, save the weights, load the weight network, obtain the model's predicted output, and thus derive the predicted traffic flow. The predicted results are then input into the embedded MCU.

[0160] The specific implementation process of step (3) is as follows:

[0161] ① Install a laser speed meter at the edge of the road at one end of the speed bump 40, and the laser outlet of the laser rangefinder is flush with the ground. The principle is to generate a narrow laser beam through the internal laser transmitter, and emit it to the vehicle being tested that is passing through the speed bump 40, and measure the time it takes for the laser beam to be reflected by the vehicle surface and then return to the speed meter. This time is the flight time t of light. According to the constant speed of light and the time it takes for light to pass, the distance d between the vehicle being tested and the laser speed meter can be obtained.

[0162]

[0163] Among them, c represents the speed of light, which is 299792458 m / s in a vacuum; n represents the atmospheric refraction coefficient, which is about 1.0003;

[0164] During the speed measurement process, the laser speed meter will perform two laser ranging operations and record the time interval as the time difference T. By recording the distance moved by the measured vehicle during these two ranging processes, the average speed v of the vehicle during this period can be calculated.

[0165]

[0166] Where d1 and d2 represent the distance between the car and the speedometer during the first and second distance measurements respectively;

[0167] After obtaining the vehicle speed data, input it into the embedded microcontroller;

[0168] ② Determine the traffic status based on the vehicle speed data. The traffic status of the lane during the green light period:

[0169] The speed v of the vehicle when passing through the speed bump 40 is close to or slightly lower than the speed limit value specified by the road, and it is judged that the lane is normal;

[0170] The speed v of the vehicle passing through the speed bump 40 is far lower than the speed limit specified by the road, and it is judged that the lane is blocked;

[0171] When the vehicle passes through the speed bump 40, the speed v is 0 for a long time and the stagnation time exceeds 15 minutes, which is judged as a lane accident;

[0172] Traffic operation status of the road in a traffic light cycle:

[0173] If all three lanes are normal, the road is considered normal;

[0174] If at least two of the three lanes are blocked, the road is considered blocked;

[0175] If one of the three lanes has an accident, it is considered a road accident;

[0176] ③ Alarm through the monitored traffic operation status. The embedded single chip computer transmits the monitored road traffic operation status to the corresponding traffic light through real-time data exchange and analysis to optimize the traffic light control and improve the road traffic efficiency. At the same time, it also transmits it to the wireless communication module. The wireless communication module uploads the road traffic operation status to the Internet for data sharing, realizes the sharing of road traffic information, and has an automatic alarm function. When the wireless communication module identifies a road accident, it will send the alarm information of the road accident to the transportation department in a timely and effective manner.

[0177] The specific implementation process of step (4) is as follows:

[0178] ① Based on the collected traffic flow data, the traffic light duration for each time period is pre-allocated. The traffic light assembly 41 is located in a two-way three-lane road, and the four traffic phases are the first phase: east→west, west→east, the second phase: east→south, west→north, the third phase: north→south, south→north, the fourth phase: south→west, north→east, and the green light time of the first phase is S T1

[0179]

[0180] Among them, max(V x , V d ) represents the traffic flow in the 1-minute time window of the west-bound straight lane and the traffic flow in the 1-minute time window of the east-bound straight lane during the historical period of 7:30-9:30 from Monday to Friday, the maximum traffic flow, L is the distance from the speed bump 40 to the zebra crossing, v a is the maximum speed of the west-bound through lane and the east-bound through lane during this time period, and t0 is the buffer time, usually 1s;

[0181] The second phase green light time S T2

[0182]

[0183] Among them, max(V X , V D) represents the traffic flow in the 1-minute time window of the west-bound left-turn lane and the traffic flow in the 1-minute time window of the east-bound left-turn lane during the historical period of 7:30-9:30 from Monday to Friday, the maximum traffic flow, L is the distance from the speed bump 40 to the zebra crossing, v b is the maximum speed of the west-bound left-turn lane and the east-bound left-turn lane during this time period, and t0 is the buffer time, usually 1s;

[0184] The third phase green light time S T3

[0185]

[0186] Among them, max(V b , V n ) represents the traffic flow in the 1-minute time window of the north-bound through lane and the traffic flow in the 1-minute time window of the south-bound through lane during the historical period of 7:30-9:30 from Monday to Friday, the maximum traffic flow, L is the distance from the speed bump 40 to the zebra crossing, v c is the maximum speed of the northbound through lane and the southbound through lane during this time period, and t0 is the buffer time, usually 1s;

[0187] Fourth phase green light time S T4

[0188]

[0189] Among them, max(V B , V N ) represents the traffic flow in the 1-minute window of the north-bound left-turn lane and the traffic flow in the 1-minute window of the south-bound left-turn lane during the historical period of 7:30-9:30 from Monday to Friday, the maximum traffic flow, L is the distance from the speed bump 40 to the zebra crossing, v c is the maximum speed of the northbound left-turn lane and the southbound left-turn lane during this time period, and t0 is the buffer time, usually 1s;

[0190] The first phase red light time S t1 =S T2 +S T3 +S T4 +9;

[0191] The second phase red light time S t2 =S T1 +S T3 +S T4 +9;

[0192] The third phase red light time S t3 =S T1 +ST2 +S T4 +9;

[0193] Fourth phase red light time S t4 =S T1 +S T2 +S T3 +9;

[0194] The total cycle of the traffic light in this time period is S T1 +S T2 +S T3 +S T4 +12, where each phase change is separated by a 3s yellow light;

[0195] ② Calculate the corresponding short-term traffic light duration based on the short-term traffic flow data predicted by LSTM, and compare it with the traffic light duration in the current time period. If there is a large gap, adjust the traffic light duration in the current time period according to the short-term traffic light duration. If there is only a small gap, do not make any adjustment and maintain the traffic light duration in the current time period.

[0196] Calculation of short-term traffic light time in the predicted future time period, first phase green light time ω T1

[0197]

[0198] in, Indicates the selected predicted future time period, the traffic flow of the 1-minute time window of the west-bound straight lane and the traffic flow of the 1-minute time window of the east-bound straight lane, the maximum traffic flow, L is the distance from the speed bump 40 to the zebra crossing, v c is the maximum speed of the west-bound through lane and the east-bound through lane in the future time period, which is determined by The obtained,t0 is the buffer time, usually 1s;

[0199] Set the first phase green light time ω in the future time period T1 The first phase green light time S in the current time period T1 For comparison, set the difference threshold to 2s. T1 -S T1 |>2, then according to ω T1 Replace S with the value T1 The value of T1 -S T1 |≤2, then S is still maintained T1 By predicting the short-term traffic flow data in the future time period, the second phase green light time ω of the future time period is calculated. T2 , and the second phase green light time S in the current time period T2Compare and calculate the second phase green light time ω in the future time period in turn T3 , compare the duration of the traffic lights in the four phases of the future time period with the duration of the traffic lights in the four phases of the current time period. If the difference exceeds the threshold, the duration of the traffic lights is adjusted. If the difference does not exceed the threshold, the duration of the traffic lights is not adjusted and the status quo is maintained;

[0200] The predicted short-term traffic flow data will be presented in chronological order and substituted into the calculation to continuously determine whether the traffic light duration needs to be adjusted. When the adjustment is made, the traffic light duration in the future time period will replace the traffic light duration in the current time period to complete the first regulation;

[0201] ③ Through real-time monitoring of the three traffic operation states of normal lane, lane congestion, and lane accident, set a second control strategy for the duration of traffic lights in different states;

[0202] When the lane is normal, keep the traffic light duration after the first adjustment to ensure smooth traffic flow;

[0203] In the case of lane congestion, determine the phase corresponding to the lane, extend the green light time of the phase after the first adjustment by 2s, and shorten the red light time by 2s, that is, shorten the green light time of other phases without lane congestion after the first adjustment by an average of 2s in total. If lanes in all four phases are congested, only the green light time will be extended and the red light time will not be shortened. When multiple lanes are congested, the lanes to which the phase belongs have If a lane is blocked, the green light time after the first adjustment of this phase will be extended Shortened red light time

[0204] In the event of a lane accident, the phase corresponding to the lane is determined, and the green light time after the first adjustment of that phase is set to 0, and the red light time is extended indefinitely and keeps flashing in an alarm state until the alarm is lifted. The traffic light duration after the first adjustment of that phase can be restored to normal to avoid further traffic congestion and accident risks.

[0205] The speed bump 40 of the road intelligent traffic multifunctional monitoring and optimization system of the present invention is provided with a gear generator 22. When a vehicle passes through the speed bump 40, it rolls over the upper cover 7, causing the upper cover 7 to move downward, thereby causing the rack 10 to drive the gear generator 22 to generate electricity through gear transmission, and is provided with a hydraulic generator 26. When a vehicle passes through the speed bump 40, the liquid storage bag 25 is squeezed, and the hydraulic oil enters the hydraulic generator 26 through the first circulation pipeline, converting the pressure of the hydraulic oil into electrical energy, and the hydraulic oil enters the liquid storage tank 27 through the second circulation pipeline for short-term storage. When no vehicle passes, the liquid storage bag 25 is reset under the action of the second reset spring, so that the hydraulic oil flows from the liquid storage tank 27 through the second circulation pipeline and the first circulation pipeline back to the liquid storage bag 25 for secondary power generation. While reducing the speed of the vehicle, it also realizes the reuse and conservation of energy, which can reduce the dependence on traditional energy and reduce the energy cost of the city.

[0206] The road intelligent traffic multifunctional monitoring and optimization system of the present invention not only has the ability to detect and predict traffic flow in real time, but also has the function of monitoring road traffic operation conditions and issuing alarms, and improves the travel experience of urban residents by intelligently regulating traffic flow.

[0207] In summary, the present invention can intelligently control traffic conditions, improve the travel experience of urban residents, and effectively solve the technical problem that the existing road intelligent traffic multifunctional monitoring and optimization system and method cannot intelligently control traffic conditions, resulting in the traffic light time being cycled within a fixed time interval, further leading to unnecessary waiting time or traffic congestion in some cases.

[0208] The above embodiments are only used to illustrate rather than limit the technical solutions of the present invention. Although the present invention is described in detail with reference to the above embodiments, those skilled in the art should understand that the present invention can still be modified or replaced by equivalents. Any modification or partial replacement that does not depart from the spirit and scope of the present invention should be included in the scope of the claims of the present invention.

Claims

1. The multifunctional monitoring and optimization system for intelligent road traffic is characterized by: It includes a controller, a speed bump and a traffic light assembly. The controller includes an embedded single chip microcomputer, a vehicle flow detection module, a short-term vehicle flow prediction module, a traffic status analysis module and a real-time traffic flow control adjustment module; The speed bump includes a left round head unit and a right round head unit which are arranged opposite to each other, and a plurality of interconnected middle units which are arranged between the left round head unit and the right round head unit; The intermediate unit includes a base plate, a front block, a rear block, an upper cover, a hydraulic power generation mechanism, and a gear power generation mechanism. The front block and the rear block are respectively arranged at the front and rear ends of the base plate, and the hydraulic power generation mechanism and the gear power generation mechanism are both arranged on the base plate and located between the front block and the rear block. The upper cover body includes a cover plate portion and a cylindrical portion, the upper surface of the cover plate portion is set as an arc surface, the lower surface of the cover plate portion is set as a plane, the upper end of the cylindrical portion is connected to the lower surface of the cover plate portion, the outer contour surface of the cylindrical portion is provided with a left-right symmetrical rack portion, a plurality of pressure power generation sheets are provided on the lower surface of the cylindrical portion, and a mounting hole is opened at the center position of the lower surface of the cylindrical portion; when a vehicle rolls over the upper cover body, pressure is applied to the pressure power generation sheet, so that the pressure power generation sheet generates electricity; The base plate is provided with a mounting column corresponding to the mounting hole and a first return spring is provided on the mounting column, and the base plate and the upper cover body are assembled and installed by matching the mounting column and the mounting hole; there are two gear power generation mechanisms, which are respectively arranged on the left and right sides of the mounting column, and the gear power generation mechanism includes a primary gear shaft, a secondary gear shaft, a primary large gear, a secondary large gear, a primary small gear, a secondary small gear, a power generation gear, a gear generator, a front vertical plate, and a rear vertical plate; the front vertical plate and the rear vertical plate are arranged opposite to each other, and the front vertical plate and the rear vertical plate are provided with corresponding first upper holes, second upper holes, first lower holes, and second lower holes, and the primary gear shaft is installed on the first upper hole and the second upper hole through the first upper hole and the second upper hole The secondary gear shaft is installed at the upper part between the front vertical plate and the rear vertical plate through the first lower hole and the second lower hole, the first large gear is installed at the front end of the first large gear shaft, the first small gear is installed at the rear end of the first large gear shaft, the first small gear is meshed with the rack part, the second large gear is installed at the rear end of the second large gear shaft, the second small gear is installed at the front end of the second large gear shaft, the second small gear is meshed with the first large gear, the second large gear is meshed with the power generation gear, and the power generation gear is set on the output shaft of the gear generator through a key connection; the vehicle passes through the speed reduction belt, rolling over the upper cover body, so that the upper cover body moves downward, and then the rack part drives the gear generator to generate electricity through the gear transmission; The electrical output terminal of the speed bump is connected to the traffic light assembly and the power storage box through electric wires, and the power storage box is connected to the traffic light assembly; the traffic light assembly includes a traffic light base, traffic light beads, and a moving wheel, the moving wheel is arranged at the bottom of the traffic light base, a traffic light frame is installed on the traffic light base, and the traffic light beads are installed on the traffic light frame.

2. The multifunctional monitoring and optimization system for intelligent road traffic according to claim 1 is characterized in that: The hydraulic power generation mechanism includes a liquid storage bag, hydraulic oil, a hydraulic generator and a liquid storage tank; the liquid storage bag is a ring-shaped liquid storage bag, which is filled with hydraulic oil. A circulation port is arranged on the side of the liquid storage bag, and the bag is connected to the hydraulic generator through a first circulation pipe. The hydraulic generator is connected to the liquid storage tank through a second circulation pipe. A plurality of second return springs are arranged in the liquid storage bag. When a vehicle passes through a speed bump, the liquid storage bag is squeezed, and the hydraulic oil enters the hydraulic generator through the first circulation pipe, converting the pressure of the hydraulic oil into electrical energy. The hydraulic oil enters the liquid storage tank through the second circulation pipe for short-term storage. When no vehicle passes by, the liquid storage bag is reset under the action of the second return spring, so that the hydraulic oil flows back from the liquid storage tank through the second circulation pipe and the first circulation pipe into the liquid storage bag for secondary power generation.

3. The multifunctional monitoring and optimization system for intelligent road traffic according to claim 2 is characterized in that: A first left pull ring and a first right pull ring are respectively arranged on the left and right sides of the lower surface of the upper cover body, a second left pull ring corresponding to the first left pull ring and a second right pull ring corresponding to the first right pull ring are arranged on the base plate, a first tension spring is installed between the first left pull ring and the second left pull ring, and a second tension spring is arranged between the first right pull ring and the second right pull ring.

4. The multifunctional monitoring and optimization system for intelligent road traffic according to claim 3 is characterized in that: A plurality of grooves arranged front to back are provided on the lower surfaces of the left round head unit, the middle unit and the right round head unit, and a plurality of heating wires are arranged in the bottom of the left round head unit, the middle unit and the right round head unit. When the heating wires are energized, the asphalt pavement is heated so that the uneven parts of the pavement are melted into the grooves to eliminate the gaps.

5. The multifunctional monitoring and optimization system for intelligent road traffic according to claim 4 is characterized in that: A first drainage ditch arranged on the left and right is provided at the connection between the base plate and the front block, a second drainage ditch arranged on the left and right is provided at the connection between the base plate and the rear block, a first drainage ditch connected to the first drainage ditch is provided on both the left and right sides of the front block, and a second drainage ditch connected to the second drainage ditch is provided on both the left and right sides of the rear block; A long strip block is arranged on the mounting column, and the mounting holes are arranged with long strip holes at the front and rear, and a rotating hole is arranged in the cylindrical part, so that the long strip block enters the long strip hole, rotates and is supported, and the first return spring is supported by the lower surface of the cylindrical part, so that when the vehicle passes through the speed bump, the upper cover body compresses the first return spring downward, and the long strip block moves upward in the rotating hole; The left round head unit and the right round head unit have the same structure, both comprising a connecting portion and a round head portion, wherein the connecting portion is arranged between the round head portion and the middle unit.

6. A road intelligent traffic multifunctional monitoring and optimization method, implemented based on the road intelligent traffic multifunctional monitoring and optimization system as claimed in any one of claims 1 to 5, characterized in that: The steps include: (1) Real-time monitoring of vehicle flow and type on the road; (2) Predict short-term road vehicle flow based on LSTM; (3) Analyze road traffic conditions; (4) Make real-time intelligent adjustments to the duration of traffic lights to achieve intelligent adjustments for real-time traffic flow control.

7. The road intelligent traffic multifunctional monitoring optimization method according to claim 6 is characterized by: The specific implementation process of step (1) is as follows: ①Convert the pressure signal generated by the vehicle passing through the speed bump into an electrical signal; ② Set different thresholds to distinguish different types of vehicles passing through speed bumps; ③ Count and record the number of vehicles passing through the speed bump.

8. The multifunctional monitoring and optimization method for intelligent road traffic according to claim 7 is characterized in that: The specific implementation process of step (2) is as follows: ① Preprocess the data by collecting historical traffic flow data of speed bumps on the road and the corresponding weather and holidays, setting the input variables as historical traffic flow, weather, and holidays, and the output variable as short-term traffic flow. The data is organized into the format of X = (X(1), X(2)...X(n)), where X represents the organized data set and n represents the number of samples; The threshold method is used to process abnormal data. The data within the threshold is not deleted, and the deleted data is filled with the time series method. The threshold method is expressed as follows: 0≤X(j)≤θ, where X(j) represents the time series of traffic flow parameters at time j, which can refer to the traffic volume, weather, and holidays at time j; θ represents the threshold; Traffic flow data is filtered according to weather records, where Vi represents the traffic volume under different weather conditions, Z represents the total traffic volume of the month, i represents different weather conditions, such as sunny, cloudy, light rain, moderate rain, heavy rain, snow, and fog, and C represents the traffic volume under different weather conditions. i Represents the influence coefficient of different weather conditions, θ i Represents the weather impact intensity factor, and sets the weather threshold to snowy and foggy daysθ i The value of . Traffic flow data is filtered according to weather records, where E r represents the traffic flow under different holidays, θ r Represents the holiday impact coefficient, r is the same as the holiday, such as Spring Festival, Qingming Festival, Labor Day, Dragon Boat Festival, Mid-Autumn Festival, National Day, set the holiday threshold to Spring Festival θ r Value If the data collected by the sensor at this moment exceeds the threshold, it is considered as erroneous data and is eliminated using the threshold method; Use time series methods to handle missing data. Among them, α represents the number of data groups selected by the time series method, which is generally 2 or 3. The missing data can be supplemented by the time series method within a reasonable range; ② Train the LSTM-based short-term traffic flow prediction model: divide the preprocessed traffic flow time series data set into two parts: a training set and a test set. The training set is used for modeling, and the test set is used to evaluate the model effect. Input the training set data into the model to train the model, and learn the characteristics of the data through the forget gate, input gate, and output gate of the LSTM, mainly learning the relationship between time. f t =σ(W f b t +U f h t-1 +b f ) (4) I t =σ(W I b t +U I h t-1 +b I ) (5) The t =σ(W o b t +U o h t-1 +b o ) (6) Initialize all weight matrices W, U and bias b, and perform forward propagation for each input training sample according to the formula to gradually calculate the forget gate f t , Input Gate I t , output gate O t The output values ​​of these gates are then used to calculate the new cell state S t , and substitute the calculated output value h t . S t =f t S t-1 +I t (tanh(W s h t-1 +U s h t-1 )+b s ) (7) h t =O t tanhS t (8) Among them, f t ,I t , O t are the output values ​​of the forget gate, input gate and output gate respectively, h t-1 is the hidden state vector of the LSTM unit at the previous moment; β t is the input vector of the LSTM unit at the current moment, is the historical traffic flow, weather and holiday characteristics; σ is the sigmoid activation function, W is the weight between the connected neurons, and U is the hidden state h at the previous moment. t-1 The weight matrix of the neuron, b is the bias of the neuron; S t is the new unit state vector obtained under the control of the forget gate and the update gate, combining the memory of the previous moment and the input data of this moment, t is the final output value under the control of the output gate; The output value h is obtained by forward propagation of the LSTM model t , there needs to be a label y i Representing the real traffic data to calculate the loss, the loss function MAE formula is as follows: in, The average absolute error between the predicted and actual traffic volume is calculated, and then the weights and biases of the network are updated through back propagation. After a certain number of iterations, the loss converges and the training ends. ③ Prediction performance evaluation and result analysis: Input the historical traffic flow, weather, holidays and other features in the test set into the loaded short-term traffic flow prediction model based on the LSTM neural network, update the model's weights to adapt to the data characteristics in the test set, save the weights, load the weight network, obtain the model's predicted output, and thus derive the predicted traffic flow. The predicted results are then input into the embedded MCU.

9. The road intelligent traffic multifunctional monitoring optimization method according to claim 8 is characterized in that: The specific implementation process of step (3) is as follows: ① Install a laser speed meter at the edge of the road at one end of the speed bump, and the laser outlet of the laser rangefinder is flush with the ground. The principle is to generate a narrow laser beam through the internal laser transmitter, and emit it to the vehicle being tested that is passing through the speed bump, and measure the time it takes for the laser beam to be reflected by the vehicle surface and then return to the speed meter. This time is the flight time t of light. According to the constant speed of light and the time it takes for light to pass, the distance d between the vehicle being tested and the laser speed meter can be obtained. Among them, c represents the speed of light, which is 299792458 m / s in a vacuum; n represents the atmospheric refraction coefficient, which is about 1.0003; During the speed measurement process, the laser speed meter will perform two laser ranging operations and record the time interval as the time difference T. By recording the distance moved by the measured vehicle during these two ranging processes, the average speed v of the vehicle during this period can be calculated. Where d1 and d2 represent the distance between the car and the speedometer during the first and second distance measurements respectively; After obtaining the vehicle speed data, input it into the embedded microcontroller; ② Determine the traffic status based on the vehicle speed data. The traffic status of the lane during the green light period: When the vehicle passes through the speed bump, the speed v is close to or slightly lower than the speed limit specified by the road, and the lane is judged to be normal; The speed v of the vehicle passing through the speed bump is far lower than the speed limit specified by the road, which is judged as lane congestion; When the speed v of the vehicle passing through the speed bump is 0 for a long time and the stagnation time exceeds 15 minutes, it is judged as a lane accident; Traffic operation status of the road in a traffic light cycle: If all three lanes are normal, the road is considered normal; If at least two of the three lanes are blocked, the road is considered blocked; If one of the three lanes has an accident, it is considered a road accident; ③ Alarm through the monitored traffic operation status. The embedded single chip computer transmits the monitored road traffic operation status to the corresponding traffic light through real-time data exchange and analysis to optimize the traffic light control and improve the road traffic efficiency. At the same time, it also transmits it to the wireless communication module. The wireless communication module uploads the road traffic operation status to the Internet for data sharing, realizes the sharing of road traffic information, and has an automatic alarm function. When the wireless communication module identifies a road accident, it will send the alarm information of the road accident to the transportation department in a timely and effective manner.

10. The road intelligent traffic multifunctional monitoring optimization method according to claim 9 is characterized in that: The specific implementation process of step (4) is as follows: ① Based on the collected traffic flow data, the traffic light duration for each time period is pre-allocated. The traffic light is located in a two-way three-lane road. The four traffic phases are: Phase 1: East→West, West→East; Phase 2: East→South, West→North; Phase 3: North→South, South→North; Phase 4: South→West, North→East. The green light time for the first phase is S T1 Among them, max(V x , V d ) represents the traffic flow in the 1-minute time window of the west-bound through lane and the traffic flow in the 1-minute time window of the east-bound through lane during the historical period of 7:30-9:30 from Monday to Friday, the maximum traffic flow, L is the distance from the speed reduction belt to the zebra crossing, v a is the maximum speed of the west-bound through lane and the east-bound through lane during this time period, and t0 is the buffer time, usually 1s; The second phase green light time S T2 Among them, max(V X , V D ) represents the traffic flow in the 1-minute time window of the west-bound left-turn lane and the traffic flow in the 1-minute time window of the east-bound left-turn lane during the historical period of 7:30-9:30 from Monday to Friday, the maximum traffic flow, L is the distance from the speed reduction belt to the zebra crossing, v b is the maximum speed of the west-bound left-turn lane and the east-bound left-turn lane during this time period, and t0 is the buffer time, usually 1s; The third phase green light time S T3 Among them, max(V b , V n ) represents the traffic flow in the northbound through lane and the southbound through lane within the 1-minute time window from 7:30 to 9:30 from Monday to Friday, the maximum traffic flow, L is the distance from the speed reduction belt to the zebra crossing, v c is the maximum speed of the north and south through lanes during this time period, t0 is the buffer time, usually 1s; Fourth phase green light time S T4 Among them, max(V B , V N ) represents the traffic flow in the 1-minute window of the north-bound left-turn lane and the traffic flow in the 1-minute window of the south-bound left-turn lane during the historical period of 7:30-9:30 from Monday to Friday, the maximum traffic flow, L is the distance from the speed reduction belt to the zebra crossing, v c is the maximum speed of the northbound left-turn lane and the southbound left-turn lane during this time period, and t0 is the buffer time, usually 1s; The first phase red light time S t1 =S T2 +S T3 +S T4 +9; The second phase red light time S t2 =S T1 +S T3 +S T4 +9; The third phase red light time S t3 =S T1 +S T2 +S T4 +9; Fourth phase red light time S t4 =S T1 +S T2 +S T3 +9; The total cycle of the traffic light in this time period is S T1 +S T2 +S T3 +S T4 +12, where each phase change is separated by a 3s yellow light; ② Calculate the corresponding short-term traffic light duration based on the short-term traffic flow data predicted by LSTM, and compare it with the traffic light duration in the current time period. If there is a large gap, adjust the traffic light duration in the current time period according to the short-term traffic light duration. If there is only a small gap, do not make any adjustment and maintain the traffic light duration in the current time period. Calculation of short-term traffic light time in the predicted future time period, first phase green light time ω T1 in, represents the traffic flow in the 1-minute time window of the west-bound straight lane and the traffic flow in the 1-minute time window of the east-bound straight lane in the selected future time period, the maximum traffic flow, L is the distance from the speed reduction belt to the zebra crossing, v c is the maximum speed of the west-bound through lane and the east-bound through lane in the future time period, which is determined by The obtained,t0 is the buffer time, usually 1s; Set the first phase green light time ω in the future time period T1 The first phase green light time s in the current time period T1 For comparison, set the difference threshold to 2s. If |ω T1 -S T1 |>2, then according to ω T1 Replace S with the value T1 The value of T2 -S T1 |≤2, then S is still maintained T1 By predicting the short-term traffic flow data in the future time period, the second phase green light time ω of the future time period is calculated. T2 , and the second phase green light time S in the current time period T2 Compare and calculate the second phase green light time ω in the future time period in turn T3 , compare the duration of the traffic lights in the four phases of the future time period with the duration of the traffic lights in the four phases of the current time period. If the difference exceeds the threshold, the duration of the traffic lights is adjusted. If the difference does not exceed the threshold, the duration of the traffic lights is not adjusted and the status quo is maintained; The predicted short-term traffic flow data will be presented in chronological order and substituted into the calculation to continuously determine whether the traffic light duration needs to be adjusted. When the adjustment is made, the traffic light duration in the future time period will replace the traffic light duration in the current time period to complete the first regulation; ③ Through real-time monitoring of the three traffic operation states of normal lane, lane congestion, and lane accident, set a second control strategy for the duration of traffic lights in different states; When the lane is normal, keep the traffic light duration after the first adjustment to ensure smooth traffic flow; In the case of lane congestion, determine the phase corresponding to the lane, extend the green light time of the phase after the first adjustment by 2s, and shorten the red light time by 2s, that is, shorten the green light time of other phases without lane congestion after the first adjustment by an average of 2s in total. If lanes in all four phases are congested, only the green light time will be extended and the red light time will not be shortened. When multiple lanes are congested, the lanes to which the phase belongs have If a lane is blocked, the green light time after the first adjustment of this phase will be extended Shortened red light time In the event of a lane accident, the phase corresponding to the lane is determined, and the green light time after the first adjustment of that phase is set to 0, and the red light time is extended indefinitely and remains in a flashing alarm state until the alarm is lifted. The traffic light duration after the first adjustment of that phase can be restored to normal, avoiding further traffic congestion and accident risks.

Images