Background: The smart grid systems with advanced cyber technologies are vulnerable to the false data injection attacks. The attacks will bypass the online state estimation monitoring, steal energy by a false dispatch in the electricity markets, and gain financial benefits. In this paper, the FDI attack case in the electricity markets of a smart grid power system is considered with micro‐grids connected. The estimated transmitted power is modified, leading to the change of electricity prices which will bring the benefit to the attacker. Further, influence on the optimization output of energy management for a Micro‐grid has been discussed. By trading online electricity with the main grid, the micro‐grids are able to maintain power balance and reduce generation cost of the whole smart grid system power system. Then different optimal generation outputs of Micro‐grid and total generation cost of the power system are also shown and discussed. Aims: The focus of this paper is to investigate the FDI attack with limited information of the system topology and dynamics, and to discuss the economic impact of such limited attacks, where the estimated transmitted power is modified and the electricity prices are correspondingly changed for benefits. Different optimal generation outputs of Micro‐grid and total generation cost of the power system are also shown and discussed. Materials & Methods: MG‐based electricity markets model of a SG power system is considered for FDI attacks. Strategy with partial information is proposed from the attack's view.In real‐time market, optimization output changes of energy management for a MG when considering the electricity trading are discussed.The quick time detection method is developed with the feature extraction technique for an accurate real‐time estimation of the fault detection, when the operator is with partial information of the system topology.Digital signal processing approaches of the discrete wavelet transform (DWT) are adopted to better extract the time‐frequency characteristics of the input transient signals and improve the detection. Results: By considering of the distributed characteristics of MGs and the optimal operation requirements of system, we transformed the problem into a standard stochastic convex program, then solved the problem by the scenario generation method. We used a modified‐IEEE standard 14‐bus system and an IEEE standard 118‐bus system to show the effect of false data injection attacks in electricity prices and the influence of the optimization output of energy management for a MG. We employed modified‐online CUSUM algorithm to establish the quickest detection of the attacks. Numerical simulation results have shown that the proposed scheme is robust to the time‐varying dynamic system, and is efficient in terms of detection accuracy for the random and structured attacks. Discussion Modified‐online CUSUM algorithm is proposed with wavelet transform and feature extraction method for the quickest detection of the attacks. Numerical simulation results have shown that the proposed scheme is robust to the time‐varying dynamic system, and is efficient in terms of detection accuracy for the random and structured attacks. The proposed detectors are also proved to have quick and reliable responses to small attack magnitudes, which are difficult to be detected by traditional CUSUM algorithm. Further, the detection method is employed on IEEE standard 118‐bus system to be compared. The simulation results have shown that the proposed model is also adapted to complex system. Conclusion: In this paper, the impact of false data injection attacks on real‐time electricity markets of the SG power system with MGs connected is investigated. Attack strategies are designed of worst‐case robustness about grid dynamics when the attacker has partial information about the grid, where the attacker can still manipulate nodal prices of the real‐time markets and maximize the profit under different degrees of uncertainties. The optimal generation outputs of MGs are discussed for the ability of maintaining power balance of the whole SG power system by trading online electricity. Besides, the detection problem with partial information of the grid topology is also investigated from grid operator's view, and a new modified detection method is proposed to defend FDI attack when considering the existence of the unknown. Modified‐online CUSUM algorithm is proposed with wavelet transform and feature extraction method for the quickest detection of the attacks. Numerical simulation results have shown that the proposed scheme is robust to the time‐varying dynamic system, and is efficient in terms of detection accuracy for the random and structured attacks. [ABSTRACT FROM AUTHOR]