At 9 o’clock, Zhu Liyu, who lives in Longchang Road, Yangpu District, drove from Zhoujiazui Road ramp to the Inner Ring Road and was drowned in the traffic that did not see the beginning and end of the morning peak. 40 minutes later, he arrived at the company parking garage on Lilac Road in Pudong, five minutes ahead of the 45 minutes predicted by the navigation software.

　　"The traffic congestion in Shanghai has indeed improved." Zhu Liyu’s experience is not accidental. A few days ago, the Traffic Analysis Report of Major Cities in China in Q2 2019, which was jointly released by several institutions, announced the congestion ranking of 50 cities in the second quarter of 2019, and Shanghai ranked 15th, dropping out of the "Top Ten Blocked Cities". In the same period of last year, Shanghai ranked eighth in this list, while in the same period of 2014, the congestion situation in Shanghai ranked first.

　　What changes have taken place in Shanghai’s traffic over the years?

　　"Beauty of Shanghai Order" friends circle swiped the screen

　　Relying on scientific and technological innovation, Shanghai Public Security Bureau constantly empowers the "electronic police". Behind every picture that embodies the beauty of order, there is an "electronic policeman" who works 24 hours a day.

　　Not long ago, a group of posts "This is the beauty of Shanghai order" were screened in the WeChat circle of friends. In these pictures taken by the surveillance video, at the place where the two lanes of the elevated ramp "merge into one", vehicles pass alternately in turn, such as the same smooth zipper. Not far away, the traffic police set up a reminder sign, and the ground was also marked with the obvious words "alternate traffic". Netizens have praised it.

　　A few years ago, the phenomenon of illegal fighting for roads and grabbing lines at the "combination of two and one" on the elevated ramp was everywhere, causing artificial congestion. Such large-scale and wide-ranging illegal acts exist widely, so it is necessary to find out in time when obtaining evidence. The elevated detachment of Shanghai Public Security Traffic Police can only divert traffic in the morning and evening rush hours, and then randomly carry out special rectification at various illegal-prone turnouts, with limited effect.

　　Behind every picture that embodies the beauty of order, there is an "electronic policeman" who works 24 hours a day. In recent years, relying on scientific and technological innovation, Shanghai Public Security Bureau has continuously empowered the "electronic police". Today’s electronic police can not only collect evidence for illegal acts such as speeding, running red lights and illegal parking, but also locate and capture many traffic violations such as motor vehicle horn, not wearing seat belts, changing lanes and traffic jams, making big turns and turning small. At the same time, the original "electronic police" has been continuously upgraded, and a set of equipment can simultaneously capture a variety of common traffic violations.

　　"Clearing traffic jams and ensuring traffic smoothness" was once a high-frequency word mentioned in Shanghai traffic management. In recent years, "strictly controlling traffic violations" has gradually become the focus of Shanghai public security traffic management. Data analysis shows that traffic violations are prone to occur frequently, which is the main reason for the chaos of road traffic order in Shanghai. In the eyes of professionals, the traffic police who have the power to enforce the law will focus on the rectification of traffic violations, use the "electronic police" to efficiently collect evidence of traffic violations, and seize the "bull nose" to control traffic congestion.

　　The traffic in mega-cities is like the human circulatory system, and one place is blocked and affects the whole body. As long as there is a car accident or breaks down, there will be a long queue behind. With the strict control of traffic violations, the number of road traffic accidents, deaths and injuries in Shanghai continued to decline, down by 8.7%, 13.2% and 2.6% respectively.

　　Smart "traffic lights" can predict traffic flow.

　　Last year, a brand-new "intelligent signal lamp system" was tried out in Shanghai. This set of traffic lights can sense and collect traffic data through multiple channels, and give the best scheme of signal control and traffic organization.

　　In the past, a typical intersection was like this: there were auxiliary police on all sides to help manage non-motor vehicles and pedestrians, while the traffic police stood in the middle of the road to direct and divert traffic, and they had to run around from time to time to enforce the law.

　　Traffic guidance depends on what the traffic police see and hear. Moreover, only the guidance on the "point" can only transmit the traffic pressure at one intersection to the next intersection. Improving road traffic efficiency must be a "systematic project" of a line and a network. On the eve of the first China International Import Expo last year, a brand-new "intelligent signal lamp system" was tried out in Shanghai. This set of signal lights does not simply indicate stopping or passing, but senses and collects traffic data through multiple channels, and gives the best scheme of signal control and traffic organization.

　　Wang Liang, deputy head of the Traffic Police Corps of the Municipal Public Security Bureau, said that the data collected by this intelligent transportation system can provide a basis for the on-site management of traffic police. "For example, if you know which road and which direction the traffic load is too large, the traffic police can guide the traffic diversion in advance, dynamically allocate the right of way and improve the traffic efficiency." According to calculation, through such dynamic and timely management measures, the regional road traffic efficiency has been improved by more than 10%. The reporter learned from the Traffic Police Corps of the Municipal Public Security Bureau that before the end of this year, there will be 1,000 intersections in the city with this set of traffic lights.

　　"I thought that the road construction would be very blocked, but I didn’t expect it to be smooth." Mr. Zeng, who lives in Wuning Road, was once worried that the road closure would affect his travel because of the reconstruction of the road at his doorstep. Unexpectedly, after the construction began, temporary traffic lanes were newly built and variable lanes were set. He did not feel that the road occupation construction brought too much impact. "The refinement of Shanghai is reflected in these details."

　　"We insist on the second-by-second adjustment of signal lights, the improvement of traffic organization one by one, and the improvement of road traffic efficiency through minor changes." Wang Liang said that taking the 104 down ramps with hidden dangers of "poor downlink" in the city as an example, every traffic police detachment should go to the local down ramps for on-the-spot investigation, investigate the reasons for poor downlink and formulate solutions one by one.

　　"Yan’ an elevated Tibet South Road has a large traffic volume and is often congested. After investigation, we found that there is a blank period between the traffic lights at the two intersections before and after the ramp, and the road resources are relatively rich." After repeated research, the elevated detachment of the Traffic Police Corps of the Municipal Public Security Bureau and the local traffic police, through fine-tuning the green light signal time, used this short time to quickly release 50 to 100 vehicles. In this way, the congestion on the off-ramp can be alleviated.

　　Remediation of traffic violations of non-motor vehicles and pedestrians

　　The rectification of non-motor vehicles and pedestrians’ violation of the law is large in quantity and wide in scope, and there is a lack of management hands like motor vehicle driver’s license and license plate. Since March 24 this year, the Shanghai Public Security Bureau has launched a centralized and unified operation every week.

　　"Shanghai’s traffic rectification has been carried out for more than three years, and citizens generally report that motor vehicles are more disciplined. However, traffic violations of non-motor vehicles and pedestrians are still high, which undermines traffic order and brings many security risks. " Wang Liang said that while keeping the "three chaos and one rebellion" of motor vehicles prominent in the strict investigation of traffic violations, since last year, the traffic police department has made traffic violations of non-motor vehicles and pedestrians the focus of rectification.

　　Speaking of rectifying non-motor vehicles and pedestrians, many front-line traffic police are puzzled: on the one hand, such violations are extensive; On the other hand, there is a lack of management grips like motor vehicle driving license and license plate. Wang Liang said: "We ask the traffic police to enforce the law as strictly as non-motor vehicles and pedestrians." Since March 24 this year, Shanghai public security organs have organized weekly centralized and unified actions to tackle and rectify traffic violations of non-motor vehicles and pedestrians, with the participation of the whole police, to tackle and rectify various traffic violations of non-motor vehicles and pedestrians.

　　Police-enterprise cooperation, social governance, alleviating parking difficulties

　　In response to the needs of residents for parking at night and picking up and dropping off students, the traffic police department, together with relevant local governments and functional departments, has taken measures to alleviate the "long-standing problem" of parking difficulties.

　　When Ms. Zhang, a citizen, was stopped for crossing the road at a red light for the first time, she was not fined. The traffic police only gave her oral education after registering her identity information. This is a "progressive" law enforcement mode adopted by Shanghai traffic police against pedestrian traffic violations: the first education, the second warning, and the third and above penalties.

　　Lifestyle changes bring new challenges to traffic management. In recent years, the development of the take-away and express delivery industry, which generally relies on non-motor vehicles to solve the "last mile", has aggravated the traffic violations of non-motor vehicles. On July 1st, the off-site enforcement system of electric bicycle traffic violations based on RFID technology was put into use. At present, the construction and installation of RFID collection equipment have been completed at key intersections and sections of the outer ring road of the city, and 54,000 electric bicycles engaged in express delivery and take-away industries have been replaced with new electronic license plates. "In the past, we lacked an effective grasp of managing non-motor vehicles, and electronic license plates provided us with a solution to the problem." According to the relevant person in charge of the traffic police department of the Municipal Public Security Bureau, as of July 28, this system has seized more than 9,400 non-motor vehicle traffic violations.

　　A few days ago, the reporter saw in the Huaihai Road business circle that the sign of "temporary parking spot for non-motor vehicles of take-away enterprises" was hung in the open space around buildings such as Times Square and Hong Kong Plaza in Shanghai. From time to time, take-away riders rode their cars here to run and deliver meals. Correspondingly, the phenomenon that non-motor vehicles are parked and placed on the road is reduced.

　　Huaihai Road Business Circle gathers commercial buildings and commercial shops, and the demand for logistics and take-away services is amazing. However, Huaihai Middle Road is a non-road section, and non-motor vehicles are prohibited from passing through the whole line. Some take-away and express delivery practitioners not only ride illegally, but also stop and put them in disorder. While contacting the property to implement the parking resources, the traffic police department promotes the cooperation between the police and enterprises and social co-governance, and innovates safety management measures such as big data screening of the problem network car, digital management of bike-sharing, "one person, one car, one license and one yard" for take-away riders and personnel score management. Last week, 28 express delivery companies jointly signed a joint commitment on rider traffic safety.

　　Parking difficulty is a "long-standing problem" in traffic management. Since the traffic rectification, the city has drawn a total of 239 "yellow solid lines" roads, and parking is prohibited across the whole line, which makes many drivers feel "very inconvenient". "In addition to strict law enforcement, we also consider the actual needs." Wang Liang said that in order to alleviate the difficulty of getting on and off taxis, the traffic police department has set up more than 3,000 taxi pick-up and drop-off points at bus stops. In view of the practical needs of residents such as parking at night and picking up and dropping off students, the traffic police department actively cooperates with relevant local governments and functional departments to take measures to alleviate the parking problem. In addition to adding time-limited road parking lots and night-time road parking lots on roads where conditions permit, we also actively tap the potential, build and add parking facilities, coordinate to turn idle land into temporary parking lots, and make use of the difference in parking demand time between residential areas and surrounding commercial buildings, hospitals, etc., and use and share parking facilities at wrong times.

　　At present, the city has more than 2,300 road parking lots with more than 90,000 berths. Among them, there are 116 night road parking lots to alleviate the parking difficulties in residential areas, which can park about 3,300 vehicles, and 273 temporary road parking lots to alleviate the parking difficulties for students, which can park about 5,900 vehicles. (Reporter Wang Xianle Jian Gongbo)

China Optics of the Light Center of the original Changguang Institute is included in the topic # Intelligent Photonics 1# Ultrafast Laser 1.

Written by Cyan (Ph.D. student, University of Bristol, UK)

Author’s recommendation

Professor Goëry Genty (University of Tampere, Finland):

Our international cooperation team published a summary article in Nature Photonics on the topic of "Machine Learning and Applications in Ultrafast Photonics". A major feature of this review is that it provides a series of basic concepts for readers (such as students) who are not familiar with related fields.

At present, people don’t know how to make the best use of artificial intelligence and machine learning to promote future research, but the application of ultrafast photon technology in controlling complex laser systems and generating customized beams has not fully exerted its potential. We have solved this problem now and summarized different machine learning methods that can be used to study and control ultrafast photonic systems.

Our goal is to make readers who are not experts in this field fully understand this review and provide them with the basic knowledge of machine learning applied to ultrafast photonics. This paper mainly summarizes the design and optimization of laser and the research and control of ultrafast dynamics (the change of light field properties).

For each case, we try to explain how machine learning technology may be applied to the actual cases of interest, the benefits they can bring and the future research direction. It also explains different algorithms in machine learning and some guidelines on how to use them, which provides a quick start for researchers who are not familiar with this field.

Artistic renderings: the application of artificial intelligence in ultrafast photonics. The figure shows how the neural network controls the properties of the incident field by seeking the optimal value in a complex multidimensional space. This concept can eventually help to generate light beams with special properties. (Source: Tampere University, Finland)

With the landing of various intelligent applications, machine learning, artificial intelligence and other professional terms that use statistical techniques and numerical algorithms to perform specific tasks have become well known to the public. In optics, the early application of machine learning mostly appears in the form of genetic algorithm, which is used for pattern recognition, image reconstruction, aberration correction or optical element design.

Figure 1. Machine learning (artistic renderings)

Source: Changchun Institute of Optics and Mechanics, Light Academic Publishing Center, New Media Working Group.

Ultrafast lasers play an important role in many fields of photonics, such as communication, material processing and biological imaging. It directly contributed to the research results of three Nobel Prize levels:

Femtosecond Laser Coherence Control (1999) [Recommended Reading]

Precision Frequency Comb (2005) [Recommended Reading]

High Power Femtosecond Pulse Amplified by Chirped Pulse (2018) [Recommended Reading]

However, the design and operation technology of ultrafast lasers have not changed much compared with those when they were first developed decades ago. With the increasing demand for the quality of light source in various new applications, the complexity of laser is increasing day by day, which means that the amount of data in the process of laser design and optimization will explode, making the traditional design optimization scheme no longer applicable.

The high-speed processing of a large amount of data is the field that AI technology such as machine learning is good at. Therefore, the combination of the latest progress of ultrafast photonics with powerful tools such as artificial intelligence and machine learning provides a new way to overcome this bottleneck.

Professor Goëry Genty from Tampere University, Finland [introduction], together with scholars from three other universities in Europe, reviewed the specific fields in which machine learning technology has been applied and breakthroughs have been made in ultrafast photonics; Consider the future challenges and research direction, and provide a feasible research process roadmap; It also points out the applications that are expected to have a significant impact in the next few years.

1. Application of machine learning in ultrafast photonics

In view of the excellent performance of machine learning in data classification, identifying hidden structures and processing parameters with large degrees of freedom, many important scientific research achievements have come out, especially in the fields of big data processing and reverse design. As shown in Figure 2, different machine learning strategies and related architectures are described, and the core concepts, implementation methods and applications in ultrafast photonics are listed.

Figure 2. Main concepts and implementation methods of machine learning in ultrafast photonics.

Source: Nature Photonics/translation: Cyan (writer)

Figure 3 describes the general application model of machine learning in ultrafast photonics. A) The diagram depicts the training stage, in which the control signal and the measuring device are used to detect the parameter space and map the corresponding operating state respectively. Then, the collected data is input into the machine learning algorithm for training.

Fig. 3. Machine learning process of controlling components with feedback loop and control algorithm.

A) training stage b) machine learning assistant stage

Source: Nature Photonics/translation: Cyan (writer)

B) Figure depicts the self-adjusting process, in which the real-time operating state characteristics of laser and a simplified measuring system are fed into electronic equipment controlled by machine learning algorithm to lock the system to an ideal mechanism. This is where machine learning is particularly powerful, because once trained, the algorithm can quickly select parameters to achieve the best operation.

When using the machine learning model, you should also pay attention to the following items:

Selection of architecture and related parameters

accuracy control

Training data

Overfitting avoidance

Robustness and Transfer Learning

According to the selection of the above structural parameters, with different algorithms in the implementation method circle in Figure 2, such as genetic algorithm, feedforward neural network, convolutional/recursive neural network and unsupervised learning, the specific target design can be realized quickly.

Second, the laser design and optimization

(1) Self-tuning of ultrafast fiber laser

Since the pulse state is influenced by many factors, such as dispersion, nonlinearity and dissipative effect, the operation of laser system is actually very complicated. This complexity, on the one hand, has greatly improved the quality of light sources; On the other hand, it also poses great challenges for control and optimization.

This difficulty stems from the need to achieve the balance of multiple degrees of freedom (or control parameters) in order to achieve stable operation or achieve a specific mechanism. When multiple output characteristics need to be optimized at the same time, the traditional optimization method will no longer be applicable. The optimal design method based on machine learning is especially suitable for this system with greatly increased complexity.

(2) Coherent dynamic control

In addition to directly controlling the laser as mentioned above, external cavity modulation technology is also widely used in pulse shaping of ultrafast pulses. This kind of optimization involving multiple related parameters, machine learning can obviously surpass other forms of automatic optimization control theory. For example, Professor Daniel B. Turner’s team showed people how the adaptive neural network algorithm can control the pulse shaper and significantly accelerate its implementation, and its convergence speed is about 100 times faster than the traditional algorithm.

Third, the characteristics and control of ultrafast pulse propagation dynamics

(1) Hidden physical model

The application of machine learning to derive prediction models from sparse or noisy measurements has been successful. With the development of technology, a concept called "hidden physical model" is put forward, that is, by using "physical information neural network" to analyze dynamic data samples, the closed mathematical model or nonlinear differential equation that controls the physical system can be automatically identified. This method should also be used in ultrafast photonics to analyze the transmission dynamics of pulses in optical fibers. In addition, this model-free method is also used to predict the coherent dynamics of soliton-like transmission. However, at present, this kind of work is only based on numerical data-the next step in this field is obviously to explore the control model from the experimental data set.

(2) Chaotic system and instability

In the chaotic system, the measurement system that directly captures the breathing laser [] imposes serious restrictions on the experiment in complexity.

Machine learning can directly solve this problem. By training neural network to determine the time characteristics of chaotic field, only the intensity characteristics of spectrum need to be measured (which is easier to measure). Recently, a similar method has been used to determine the peak power, duration and time delay of extreme wandering solitons in noise continuous spectrum. In addition, similar applications have been extended to more complex systems, such as those observed in transient laser behavior and extreme events.

(3) Multidimensional system

One of the main advantages of neural network is that it can effectively analyze the characteristics of multidimensional systems. This is especially suitable for multimode optical fiber systems, because space-time coupling greatly increases the complexity of parameter space and nonlinear propagation dynamics. Uurtein of the Federal Institute of Technology in Lausanne, Switzerland, drives a spatial light modulator through a neural network to generate supercontinuum spectrum in a step index distribution fiber, which proves the potential of machine learning.

Figure 4. Neural network structure (artistic effect diagram)

Source: Changchun Institute of Optics and Mechanics, Light Academic Publishing Center, New Media Working Group.

IV. Prospects and Challenges

Ultrafast photonic systems are very complex and usually nonlinear, and their dynamic characteristics are very sensitive to internal parameters and external disturbances. Due to the demand for stability, anti-interference robustness, adjustability and adaptive control, the complexity of these systems is increasing day by day. Machine learning technology, which can find hidden features and adapt independently when contacting new data, may play an important role in the next generation of ultra-fast systems and applications.

At the same time, in the application of ultrafast photonics, machine learning also faces the following challenges:

1) When using recursive networks, it is very important to sample pairs along the evolutionary dimension (time or distance) to extract and reproduce the long-term evolutionary structure. Therefore, memory limitation will become a problem, especially in lasers that require multiple cavity round trips to stabilize.

2) Unsupervised learning analysis divides data into subsets with similarities, but it lacks key information about the criteria used for classification.

3) When the photon system is processed in real time, it needs the ability to manage a large amount of data and a hardware framework that can cope with the ultra-fast processing speed.

To sum up, machine learning provides effective solutions for many problems encountered at this stage, which makes ultrafast photonics play an effective role in a wider range of fields. As Professor Goëry Genty said: "We believe that machine learning can play an important role in the next generation of laser light sources. Our work can inspire researchers to fully tap the potential of artificial intelligence in ultrafast photonics and its many applications. "

Article information

Genty, G., Salmela, L., Dudley, J.M. et al. Machine learning and applications in ultrafast photonics. Nat. Photonics (2020).

Original address

https://doi.org/10.1038/s41566-020-00716-4

【】

Welcome the research group to submit the publicity draft of the results.

Reprinted/cooperated/contributed by the research group, please add WeChat: 447882024.

Previous recommendation

Light: Science & Applications | Noise reduction of light: suppression of backscattering of light.

Nature Photonics | "remote control" of polarized light

Progress of Science | Micro Spectrometer in the Past 30 Years

New publication recommendation

Open manuscript: Light: Advanced Manufacturing

ISSN 2689-9620

Periodical website: www.light-am.com

Opening manuscript: eLight

CN 22-1427/O4 ISSN 2662-8643

Periodical website: elight.springeropen.com

This is the 1631st article published by China Optics. If you think it is helpful, forwarding it to the circle of friends is our greatest recognition.

【】

Original title: Application of Machine Learning in Ultrafast Photonics

Read the original text