CONCEPT OF THE INTELLIGENT SYSTEM DEVELOPMENT FOR ACTIVE VEHICLE SAFETY

Transcription

1 International Journal of Civil Engineering and Technology (IJCIET) Volume 9, Issue 11, November 2018, pp , Article ID: IJCIET_09_11_256 Available online at ISSN Print: and ISSN Online: IAEME Publication Scopus Indexed CONCEPT OF THE INTELLIGENT SYSTEM DEVELOPMENT FOR ACTIVE VEHICLE SAFETY Andrei Mikhailovich Saikin, Sergei Evgenevich Buznikov and Vladislav Olegovich Strukov Central Scientific Research Automobile and Automotive Engines Institute NAMI , Moscow, Russia ABSTRACT The aim of the research is to provide a substantiation for the concept of the intelligent system development of active vehicle safety. The research methods include system analysis, multicriteria optimization and modern control theory. A list of the main criteria for assessing the consumer performance of technical control systems is given. The results of the study are presented in the form of the analysis of the properties of the modified management quality functional and associated consumer indicators. Intellectual properties are acquired by increasing the monitoring and management functions in a minimal hardware configuration. Discussion of the research results is carried out in the form of a comparative analysis of sensor and intelligent systems. The analysis of the research results allows stating that the solution of minimizing the quadratic functional of control quality and the multicriteria problem is among software and hardware solutions of intelligent systems realized in the minimal configuration of hardware. The novelty of the work is the scientific substantiation of the concept of intelligent system development of active vehicle safety based on virtual information sensors aimed at improving competitiveness, using new mathematical models and control algorithms that provide the solution of the problem of crash avoidance in the most general formulation. Key words: System Analysis, Multi-Criteria Optimization, Quadratic Functional of Control Quality, Virtual Sensors, Dynamic Stabilization Problem Cite this Article: Andrei Mikhailovich Saikin, Sergei Evgenevich Buznikov and Vladislav Olegovich Strukov, Concept of the Intelligent System Development for Active Vehicle Safety, International Journal of Civil Engineering and Technology (IJCIET) 9(11), 2018, pp INTRODUCTION The problem of road transport safety concerns the vital interests of almost all members of modern society and has a world significance level editor@iaeme.com

2 Concept of the Intelligent System Development for Active Vehicle Safety The problem of traffic safety becomes especially urgent under the conditions of the development of unmanned vehicles operated on public roads. Leading foreign [1] and Russian companies [2] carry out large-scale research in the field of intelligent transport systems [3] and unmanned vehicles. Their practice shows that the transition to unmanned control is connected not only with the development of adaptive cruise control systems [4], technical vision [5] and navigation [6]. Safe operation of unmanned vehicles is impossible without the problem of crash avoidance being solved. A promising direction, developed by leading foreign companies [7], for solving this problem is the development of effective systems of active vehicle safety. An active safety system (ASS) is supposed to prevent collisions of the vehicle with obstacles, while passive safety systems serve to reduce the severity of the consequences of a crash in case it could not be prevented. System analysis [8] of conditions for a collision makes it possible to construct a set of typical collisions [9]. The structured set of collisions contains 90 typical collisions with obstacles and 10 rollovers. The causes of such emergency events are technical malfunctions of the car, the pavement condition and driver control errors. Modern foreign ASS [10-13] make it possible under certain conditions to solve the problems of automatic stability of longitudinal (ABS and ASR) and transversal (ESP) wheel slide, speed of center-of-mass and distance (ACC) to obstacles in front hemisphere on the lane, advanced emergency braking (AEBS), lane position control (LDWS), tire pressure monitoring (TPMS), etc. A total number of potentially preventable typical collisions by such systems does not exceed 57 out of 100 possible. At the same time, it should be taken into account that equipping vehicles with full-featured active safety systems is not a guarantee against collisions with other traffic participants not equipped with such systems. This means that collision avoidance systems should be installed on all vehicles, including models of low price categories, which determines the low cost of technical means and the costs of their operation. Possibilities of almost unlimited growth of functions with minimal configuration of technical means are possessed by intelligent control systems. The mentioned properties are acquired by using mathematical models and indirect measurement algorithms, key for controlling physical variables. Physical sensors are replaced by virtual, software-programmable, which significantly reduces the cost of replication and maintenance of such systems. The concept of the development of an intelligent active safety system (IASS) includes a group of objectives for the development of IASS, the achievement of which is possible with the use of new technologies, models and methods of building information control systems. 2. METHODS To justify the concept of creating an intelligent system of active car safety, multicriterion optimization methods are used in the area of hardware (R H ) and software (R S ) tools [14], available to the developer. When developing and justifying the criteria for evaluating on-board ASS designed to prevent certain groups of typical wheeled vehicle collisions, an analysis of the main criteria for evaluating technical systems should be carried out. Summarizing the practice from working with clients - consumers of high-tech technical products in the field of automotive electronics - allows determining the main consumer indicators of interest to the end user editor@iaeme.com

3 Andrei Mikhailovich Saikin, Sergei Evgenevich Buznikov and Vladislav Olegovich Strukov As the general criteria for evaluating technical systems, the following consumer indicators can be considered, which are reducible to the minimized form: level of external design; completeness of implemented functions: level of observation and control errors; level of power consumption from external sources; degree of influence of external factors; degree of influence on adjacent systems; level of fault tolerance; restrictions on installation at sites; operating costs; price of the system kit. Requirements for the level of external design of devices, that are a part of embedded systems and the convenience of using them in the management process are decisive when they are selected from a competing variety of options. Therefore, it is one of the most important consumer indicators of technical products. The creative area of industrial design in the field of automotive electronics is unlimited, but the results of real activity remain wanting, because of objective and subjective reasons that limit the development of this area in Russia. The completeness of the functions being implemented is the most important criterion for evaluating possible technical solutions of the onboard ASS. Since the aim of the onboard ASS is to prevent typical collisions out of a structured set, then the more typical collisions are prevented by the operation of ASS, the more effective is its use and the less the number of typical situations not provided with protection. The developed system provides the implementation of all the functions of known ASS by combining the functions of measuring and actuating devices and the integrated special software. Special requirements for anti-lock braking systems, road-holding stabilization systems, automatic emergency and auxiliary braking systems, tire pressure monitoring and the lanekeeping control are formulated in the relevant sections of the UNECE Regulations. The consumer indicator characterizing the level of monitoring and control errors for onboard ASS determines the level of errors in monitoring the vehicle's state coordinates relative to their dynamic boundaries and the quality of solving the problems of control of traction, brakes and heading in real operation in the road and climate conditions of Russia. The level of power consumption from external power sources is an important evaluation criterion for ASS, taking into account the constantly increasing requirements for the economy of cars, the volume and content of harmful emissions produced by automobiles, and also considering the trends in the spread of electric vehicles. The principle of using virtual sensors, that is, algorithmic designs that allow replacing physical sensors of information, potentially allows providing a lower level of energy consumption in comparison with similar systems that use physical sensors to obtain estimates. The level of influence of external factors for ASS designed to prevent collisions is certainly an important criterion since the system must provide stable operation under different climatic conditions and possess high interference immunity. For the system being developed, editor@iaeme.com

4 Concept of the Intelligent System Development for Active Vehicle Safety these properties are achieved by using pickup gear sensors that provide reliable information in real noise conditions by appropriately selecting a threshold level of signals ensuring that no errors of the first and second kind occur in the operating speed range. The level of influence on adjacent systems is also one of the most important criteria for the evaluation of ASS since a large number of electronic systems enter the composition of a modern car, and their number and mutual integration will only increase with time, and the probability of short circuits will increase. Similarly, the ASS itself should be minimally affected by the undesirable influence of external systems, which is taken into account in the consumer indicator "the level of influence of external factors". The level of fault tolerance is also a key indicator of ASS. The use of virtual information sensors makes it possible to create information redundancy, which will be used to failure information redundancy of wheel speed sensors, which allows abandoning the expensive hardware redundancy used in such cases. It should be noted that the vast majority of modern ASS can function only in one, complete configuration of the devices included in them. Restrictions on installation for onboard ASS are the most important criterion for their applicability. Compliance with this criterion is solved by methods of standardization and unification of the technical solutions being developed. The operating costs for onboard ASS add up to the costs of diagnosing faults, maintenance, replacing and configuring faulty devices and increasing as the number of devices entering the system increases. Considering that failure-free systems cannot be reproduced until now, the need for diagnostics and maintenance remains relevant. The cost of the system kit for onboard ASS systems is one of the key criteria when choosing this variant of the technical solution among competing analogues. 3. RESULTS To provide a comparative analysis of possible IASS technical solutions, the authors use the properties of a quadratic functional of management quality and consumer indicators. At the conceptual level IASS should provide the solution of the following applied problems: to detect the approach and exceeding of the boundaries of dangerous states, performing informational support for the decision making by the driver; to automatically stabilize the longitudinal and cross slip of the wheels within safe limits; to automatically stabilize the safe speed and distance to obstacles on the lane; to perform automatic braking to a complete stop before a fixed obstacle on the lane; to automatically stabilize the course and position of the car according to the recognized markup. On a meaningful level, IASS should provide a solution to the task of dynamically stabilizing the coordinates of the state of the object. To assess the quality of the solution of the dynamic stabilization problem of a vehicle, to which the collision avoidance problem is reduced, a modified control quality functional is introduced in the final time interval of operation :, (1) where ; editor@iaeme.com

5 Andrei Mikhailovich Saikin, Sergei Evgenevich Buznikov and Vladislav Olegovich Strukov The terms of the functional characterize the quality of the solution of the problem of dynamic stabilization of the state vector within the upper and lower dynamic boundaries and. If the components of the state vector are within the upper and lower bounds, the penalty coefficients and are reset to zero, otherwise, they are assigned large positive values and. The risk functional has the property that its value is zero if such control of the U object is provided that the constraints on the state vector are satisfied. If the boundaries are violated, this means that the conditions for preventing one or more typical collisions are not met for the control object and prerequisites for an accident arise. The more time an object is in a dangerous state, the more the accumulated amount of risk becomes. The term of the control cost of the functional corresponds to the equivalent energy of the driver spent for control. The control vector components correspond to the control actions on the gearbox ( ), the throttle valve ( ), the braking system ( ), the steering wheel ( ) and the clutch pedal ( ). In the case of mechanical transmission and lack of speed stabilization systems, automatic braking and heading control, the values of the scale factors. The use of automatic transmission eliminates the need to control the driver gearbox and clutch. Accordingly, and are reset to zero. Automatic control of traction and brakes corresponds to the zeroing of and. The automatic heading control is accompanied by a zeroing of and in this case, the additive component of losses on the driver's control is also reset to zero. The term of the cost functional for the acquisition and operation of the system determines the total costs for software and hardware for the period of operation. The problem of choosing and justifying the optimal version of the onboard IASS at the content level is reduced to minimizing the modified functional due to the controls U implemented by R available in the hardware and software environment: (2) with, The best or optimal solution of the problem in the sense of the criterion (2) is the control implemented by the driver and the control system using data generated by the software and hardware environment R, providing the dynamic stabilization solution and characterized by minimal costs for purchase and operation. Consider the properties of such an optimal solution and the rationale for choosing the best variant of the software-hardware solution for IASS. To minimize the term of the risk functional, it is necessary to form the estimates of the components of the state vector and its upper and lower boundaries. For unobservable components and their bounds, the estimates can be given by constants satisfying the inequalities editor@iaeme.com

6 Concept of the Intelligent System Development for Active Vehicle Safety However, in this case, neither dangerous approximation to the boundaries nor their excess will be detected and the term of the risk functional will increase in proportion to the time of the object's stay in a dangerous state, which does not correspond to the solution of the dynamic stabilization problem. Therefore, it is necessary to provide an adequate estimate of the components of the state for all N components and their boundaries and. In case of significant measurement errors, the errors of the first and second kind are possible, i.e. false alarms and skipping the event of exceeding the boundaries. The formation of control actions by the driver with the use of information support for making decisions to prevent exceeding the permissible limits assumes the adequacy of his condition, which should also be controlled by IASS. The term of the cost functional for the acquisition and operation of software and hardware complexes: is determined by the number of the sensors indicating information about object state variables N, driver, actuators for control and software:, (3) where the cost of information sensors and actuators purchased or produced by the manufacturer; the cost of replication of special software of a developer, incommensurable with. The intensity of the costs of maintaining operating capability is also determined by the number of primary information sensors served: (4) To reduce the cost of in intelligent systems for information processing and control, virtual sensors that replace physical ones are used. Their operation is based on the use of mathematical models and algorithms of indirect measurements, which allow using the minimum number of physical information sensors and actuators. In this case, for all, inequalities are satisfied: where a technical solution with a minimum set of sensors for primary information and actuators and virtual sensors for measuring component of the object and the driver. For consumer indicators, this means that and the developed intelligent active safety system acquires competitiveness properties in a certain segment of applied technical systems. 4. DISCUSSION The trivial solution to the problem of developing IASS is to equip the object with numerous sensors of physical variables that form the conceptual core of traditional sensor control systems. This area, developed by leading foreign companies, inevitably leads to the performance degradation of almost all system indicators, including the total cost, the cost of operation, installation, maintenance, consumed energy, influence on adjacent systems, fault tolerance, etc. and requires highly qualified maintenance, possible only in the branded centers. (5) editor@iaeme.com

7 Andrei Mikhailovich Saikin, Sergei Evgenevich Buznikov and Vladislav Olegovich Strukov A non-trivial solution to the problem of developing IASS with new technologies is to use indirect measurements based on knowledge of the object in the form of formalized descriptions of its properties that form the conceptual core of intelligent systems. It should be noted that many of the problems of indirect measurements are initially incorrect and need to be redefined by the properties of the object. Intelligent systems using mathematical models and indirect measurement algorithms can be performed in a minimal hardware configuration, which allows optimizing almost all technical indicators, including cost. Table 1 shows the comparative characteristics of sensor and intelligent systems. Table 1. Characteristics of sensor and intelligent systems System characteristics Sensor systems Intelligent systems 1. Enhancement of functions Additional information sensors Additional software 2. Monitoring fault level Defined by the accuracy class Defined by the level of of sensors adequacy of models 3. Fault tolerance Hardware redundancy Information redundancy 4. Maintenance Costs Troubleshooting, replacement and adjustment of sensors Practically not required 5. Power Consumption Increases with the number of sensors Minimal, does not change 6. Influence of real factors Increases with the number of sensors Minimal 7. Influence on related systems Occurs in case of short circuits Minimal in the absence of in common power circuits common power circuits 8. Restrictions on installation Strictly determined by the design of the object Practically absent 9. Total cost Increases with the number of sensors Minimal The analysis of the results of the conducted studies allows making a conclusion that the solution of the problem of minimizing the quadratic functional of the control quality and the multicriteria problem lies in the area of software and hardware solutions of intelligent systems implemented in the minimal configuration of hardware. 5. CONCLUSIONS The concept of developing an IASS for wheeled vehicles involves achieving the following objectives: implementation of the proposed ASS in the minimal possible configuration of the elements of the component base due to the use of algorithms for indirect measurements (virtual sensors) and knowledge of the control object in the form of formalized descriptions of its properties; improving road safety by eliminating a significant part of the prerequisites for typical collisions and rollovers; reduction of costs for maintenance of vehicles due to timely detection of critical malfunctions; increasing the capacity of the road network, reducing the amount of harmful emissions to the environment by increasing the average speed of traffic flow due to a decrease in the number of accidents. Most foreign ASS for vehicles consist of a large number of physical sensors for information on the state of the car, electronic collection units, information processing and traction control, brakes and course designed to inform decision support by the driver and automatic control of the car to prevent collisions of certain types editor@iaeme.com

8 Concept of the Intelligent System Development for Active Vehicle Safety However, modern ASS, in overwhelming quantity, are complex constructively, contain an excessive number of information sensors and in the road-climatic conditions of Russia have a reduced efficiency. In addition, the studies show that modern foreign ASS potentially prevent about 57 out of 100 possible typical collisions and rollovers. Analysis of the experience, modern scientific, technical, regulatory, methodological literature on the problem of automotive ASS allows us to state that the application of mathematical models and algorithms of indirect measurements, along with more sophisticated control algorithms, are the basis for the development of efficient technologies and applied control systems in this region. The development of the experimental model of the domestic intellectual ASS and elements of the component base of the system based on virtual information sensors is focused on solving the problem of import substitution in this field. The implementation of the proposed ASS in the minimally possible configuration of the elements of the component base, through the use of algorithms and mathematical models of indirect measurements (virtual sensors) is accompanied by an abrupt improvement in almost all consumer indicators, including reliability, fault tolerance, operation costs, energy consumption and total cost. ACKNOWLEDGEMENTS This scientific article was prepared on the base of the results of applied scientific research and experimental development, which were carried out with the financial support of the State represented by the Ministry of Education and Science of Russia, Agreement No Unique identifier of the project RFMEFI62517X0042. REFERENCES [1] Rothhämel, M. Fail-operational vehicle dynamics for autonomous operation a review through other engineering domains. In Proceedings of the 25th International Symposium on Dynamics of Vehicles on Roads and Tracks (IAVSD 2017) 1, [2] Saikin, A. M., Bakhmutov, S. V., Terenchenko, A. S., Endachev, D. V., Karpukhin, K. E. and Zarubkin, V. V. Tendency of Creation of Driverless Vehicles Abroad. Biosciences Biotechnology Research Asia, 11, 2014, pp [3] Shadrin, S. S., Ivanov, A. M. and Karpukhin, K. E. Using Data from Multiplex Networks on Vehicles in Road Tests, in Intelligent Transportation Systems, and in Self-Driving Cars. Russian Engineering Research, 36(10), 2016, pp [4] Davis, L. C. Stability of adaptive cruise control systems taking account of vehicle response time and delay. Physics Letters A, 376, 2012, pp [5] Saikin, A., Buznikov, S. and Karpukhin, K. The Analysis of Technical Vision Problems Typical for Driverless Vehicles. Research Journal of Pharmaceutical, Biological and Chemical Sciences, 7(4), 2016, pp [6] Kiss, D. and Papp, D. Effective Navigation in Narrow Areas: A Planning Method for Autonomous Cars. Proceedings of the IEEE 15th Int. Symp. on Applied Machine Intelligence and Informatics (SAMI 2017) [7] Eunbi, J. Evaluating the effectiveness of active vehicle safety systems. Accident Analysis & Prevention, 100, 2017, pp [8] Ilarionov, V. A. Sistemnyi analiz dorozhnogo dvizheniya i dorozhno-transportnykh proisshestvii [System analysis of traffic and road accidents]. Moscow: MADI, editor@iaeme.com

9 Andrei Mikhailovich Saikin, Sergei Evgenevich Buznikov and Vladislav Olegovich Strukov [9] Buznikov, S. E. System analysis of vehicle active safety problem. IOP Conference Series: Materials Science and Engineering, 315, 2018, pp [10] Her, H., Yi, K., Suh, J. and Kim, C. Development of Integrated Control of Electronic Stability Control, Continuous Damping Control and Active Anti-Roll Bar for Vehicle Yaw Stability. IFAC Proceedings, 46, 2013, pp [11] Shadrin, S. and Ivanov, A. Algorithm of autonomous vehicle steering system control law estimation while the desired trajectory driving. ARPN Journal of Engineering and Applied Sciences, 11(15), 2016, pp [12] Ivanov, A. and Shadrin, S. Development of autonomous vehicles' testing system. IOP Conference Series: Materials Science and Engineering, 315(1), 2018, pp [13] Tuan, N. Kh., Karpukhin, K. E., Terenchenko, A. S. and Kolbasov, A. F. World Trends in the Development of Vehicles with Alternative Energy Sources. ARPN Journal of Engineering and Applied Sciences, 13(7), 2018, pp [14] Zwicky, F. The morphological approach to discovery, invention, research and construction. In: New methods of thought and procedure, 1967, pp