Quantum decision making in automatic driving

Abstract

The behavior intention estimation and interaction between Autonomous Vehicles (AV) and human traffic participants are the key problems in Automatic Driving System (ADS). When the classical decision theory studies implicitly assume that the behavior of human traffic participants is completely rational. However, according to the booming quantum decision theory in recent years and actual traffic cases, traffic behaviors and other human behaviors are often irrational and violate the assumptions of classical cognitive and decision theory. This paper explores the decision-making problem in the two-car game scene based on quantum decision theory and compares it with the current mainstream method of studying irrational behavior-Cumulative Prospect Theory (CPT) model. The comparative analysis proved that the Quantum Game Theory (QGT) model can explain the separation effect which the classical probability model can’t reveal, and it has more advantages than CPT model in dealing with game scene decision-making. When two cars interact with each other, the QGT model can consider the interests of both sides from the perspective of the other car. Compared with the classical probability model and CPT model, the QGT is more realistic in the behavior decision-making of ADS.

Mapping the Quantum Frontier in Decision-making: A Comprehensive Review on the Implications of QDT for Strategy and Governance

“…Unlike classical models that often marginalize the role of emotions and cognitive biases, QDT is inherently equipped to incorporate these variables into its framework (Khrennikov, 2022;Vaid & Chaudhary, 2022). The theory accommodates the complexity of human behavior, including incongruities and idiosyncrasies exhibited by individuals when making decisions (Song et al, 2022;Dervishaj & Cakerri, 2022). Furthermore, the quantum framework adopts a less deterministic approach, allowing for a wider range of potential outcomes and the possibility of change, which mirrors the true and often unpredictable nature of human decision-making (Khrennikov, 2022;Song et al, 2022).…”

“…The theory accommodates the complexity of human behavior, including incongruities and idiosyncrasies exhibited by individuals when making decisions (Song et al, 2022;Dervishaj & Cakerri, 2022). Furthermore, the quantum framework adopts a less deterministic approach, allowing for a wider range of potential outcomes and the possibility of change, which mirrors the true and often unpredictable nature of human decision-making (Khrennikov, 2022;Song et al, 2022). QDT presents a realistic model for assessing human behavior in decision-making contexts, offering a robust alternative to classical theories that may fall short in capturing the entirety of human complexity (Vaid & Chaudhary, 2022;Dervishaj & Cakerri, 2022).…”

Privacy protection of quantum BP neural network based on game theory

“…For example, Zhang et al [50] proposed a ration quantum state sharing protocol with the semi-off-line dealer based on game theory. Song et al [51] explored the decision problem in the two-car game scenario based on quantum decision theory. Existing research has combined game theory with quantum machine learning for many applications, but has not yet applied it to privacy security analysis.…”

Recognition and Prediction of Pedestrian Hazardous Crossing Intentions in Visual Field Obstruction Areas Based on IPVO-LSTM

“…As for the hazardous crossing problem caused by the irrational behavior of pedestrians, we can consider introducing the quantum-mechanics-based irrational driving behavior decision method in our previous study [49].…”